Guide to Data Sources on the Determinants of Marriage and

Guide to Data Sources on the Determinants of Marriage and Cohabitation Marriage and Family Formation Data Analysis Project Contract #282-98-0006 Cambridge, MA Lexington, MA Hadley, MA Bethesda, MD Washington, DC Chicago, IL Cairo, Egypt Johannesburg, South Africa Final Report March 2003 Prepared for Department of Health and Human Services, Administration for Children and Families, Office of Planning Research and Evaluation Abt Associates Inc. 55 Wheeler Street Cambridge, MA 02138 Prepared by Nancy R. Burstein Laura D. Lindberg David J. Fein Lindsay Page Richard J. LaRock, Jr. Acknowledgements We are grateful for guidance and support from Nancye Campbell, Naomi Goldstein, Brendan Kelly, and Howard Rolston of the Administration for Children and Families. Abt Associates staff who provided invaluable assistance were Chris Hamilton, Rebecca Burstein, and Jan Nicholson. All interpretations in this report are our own and do not necessarily reflect the views of the U.S. Department of Health and Human Services. Abt Associates Inc. Contents i Contents Acknowledgements ......................................................................................................................i Chapter One: Introduction.........................................................................................................1 Topics Covered and Units of Measurement and Analysis ....................................................2 Desirable Attributes of Data Sets.......................................................................................3 Data Sets Included in this Guide ........................................................................................6 Contents of this Guide.......................................................................................................9 Chapter Two: The Current Population Survey.........................................................................11 Survey Design ................................................................................................................ 11 Survey Content............................................................................................................... 11 Suitability for Research................................................................................................... 12 Using the Data................................................................................................................ 13 Data Elements ................................................................................................................ 14 Chapter Three: The Fragile Families and Child Wellbeing Study ............................................17 Survey Design ................................................................................................................ 17 Survey Content............................................................................................................... 18 Suitability for Research................................................................................................... 18 Using the Data................................................................................................................ 20 Data Elements ................................................................................................................ 21 Chapter Four: The National Longitudinal Survey Of Youth ....................................................33 Survey Design ................................................................................................................ 33 Survey Content............................................................................................................... 34 Suitability for Research................................................................................................... 35 Using the Data................................................................................................................ 36 Data Elements ................................................................................................................ 37 Chapter Five: The National Survey Of America’s Families ......................................................43 Survey Design ................................................................................................................ 43 Survey Content............................................................................................................... 44 Suitability for Research................................................................................................... 44 Using the Data................................................................................................................ 45 Data Elements ................................................................................................................ 46 Chapter Six: The National Survey Of Family Growth ..............................................................49 Survey Design ................................................................................................................ 49 Survey Content............................................................................................................... 51 Suitability for Research................................................................................................... 52 Using the Data................................................................................................................ 54 Data Elements ................................................................................................................ 55 Abt Associates Inc. Contents i Chapter Seven: A National Survey Of Families And Households ..............................................63 Survey Design ................................................................................................................ 63 Survey Content: Wave 1................................................................................................. 64 Survey Content: Wave 2................................................................................................. 64 Survey Content: Wave 3................................................................................................. 65 Suitability for Research................................................................................................... 65 Using the Data................................................................................................................ 67 Data Elements ................................................................................................................ 68 Chapter Eight: The Panel Study Of Income Dynamics .............................................................91 Survey Design ................................................................................................................ 91 Survey Content............................................................................................................... 91 Suitability for Research................................................................................................... 92 Using the Data................................................................................................................ 94 Data Elements ................................................................................................................ 94 Chapter Nine: The Survey Of Income And Program Participation...........................................99 Survey Design ................................................................................................................ 99 Survey Content............................................................................................................. 100 Suitability for Research................................................................................................. 101 Using the Data.............................................................................................................. 102 Data Elements .............................................................................................................. 102 Chapter Ten: The Survey Of Program Dynamics ...................................................................111 Survey Design .............................................................................................................. 111 Survey Content............................................................................................................. 112 Suitability for Research................................................................................................. 113 Using the Data.............................................................................................................. 115 Data Elements .............................................................................................................. 116 Chapter Eleven: Research Opportunities..............................................................................123 Theme 1: Teen and Non-Marital Childbearing............................................................... 123 Theme 2: Transition to Parenthood ............................................................................... 124 Theme 3: Effects of Cohabitation on Marriage ............................................................... 125 Theme 4: The Surge of Low-Income Women Into Jobs .................................................. 126 Theme 5: Men’s Economic Status ................................................................................. 127 Theme 6: The Meaning of Marriage and Cohabitation .................................................... 128 Theme 7: Gender Role Expectations ............................................................................. 130 Theme 8: Interaction Processes..................................................................................... 130 Theme 9: Intrapersonal Influences on Interaction ........................................................... 131 Theme 10: Contextual Influences on Interaction ............................................................ 132 References...............................................................................................................................135 Appendix A: Research Opportunities by Theme .....................................................................A-1 ii Contents Abt Associates Inc. Chapter One: Introduction The proportion of American children living in single parent families grew steadily from the 1960s through the mid-1990s and has remained at high levels since then (Ellwood and Jencks 2001). Declines in marriage, increases in divorce and non-marital cohabitation, and growing non-marital childbearing all have played a part in these trends. Children whose parents face limited economic opportunities are especially likely to grow up in single parent families. As a consequence, these children are at a higher risk of lagging social and emotional development, failure in school, limited career prospects, and becoming single parents themselves (McLanahan and Sandefur 1994). U.S. government policy has sought alternatively to improve the circumstances of single parent families and to reduce their numbers. There now is growing interest in interventions seeking to promote healthy marriages, and, where marriage is infeasible, strengthen relationships between unmarried parents. These interventions will stand a better chance for success if they are based on a sound understanding of the determinants of union formation, stability, and quality. The Administration for Children and Families (ACF) is the lead federal agency charged with developing and testing in itiatives in this area. Mindful of the need for strong basic empirical foundation for policies and programs, ACF commissioned Abt Associates Inc. to review the evidence on important determinants of marriage and cohabitation among disadvantaged Americans. The objectives of this project were to (1) assess findings and gaps in basic research literature and (2) provide a guide to major national survey data that could be used to close the gaps. This report— which documents nine key surveys—fulfills the second of these objectives. An accompanying report presents findings from our literature review (Fein et al. 2003). The audience for this guide is researchers and agencies who are interested in opportunities for research on particular aspects of marriage and cohabitation determinants but are not familiar with the contents of one or more of these datasets. Discerning the pertinent attributes of national surveys for a given topic can require an extensive search for documentation, and, once found, documentation for some surveys can be voluminous. We hope that the guide will help users to find their way to pertinent information more quickly. Chapters on individual surveys focus on their uses in analyses of the formation, stability and quality of marriage and cohabitation among disadvantaged populations. A concluding chapter also reviews opportunities these surveys offer to address important knowledge gaps for those topics identified in our literature review. Recognizing that the information herein is only the starting point for analysis planning, we describe how users can obtain more detailed documentation for each survey, as well as the survey data themselves. Readers may wish also to consult two useful predecessors to this guide. One is Researching the Family: A Guide to Survey and Statistical Data on U.S. Families, prepared by Child Trends Inc. in the early 1990s (Zill and Daly 1993). Although somewhat out of date, it nonetheless provides a useful summary of general features and items pertinent to a wide range of family-related analyses for 50 surveys. A more recent document is the series of data set profiles prepared by the Federal Interagency Forum on Child and Family Statistics (2001). The latter provide particularly useful documentation on measures of marriage, divorce, remarriage, and cohabitation in major federal surveys. Abt Associates Inc. Chapter 1: Introduction 1 In this introductory chapter, we start by defining the topics of analysis that this guide was designed to address. Then, we describe some of the desirable attributes of data sets for research in this area. Next, we identify the datasets included in the guide and comment on several useful datasets that we were not able to include. Finally, we outline the guide’s remaining chapters. Topics Covered and Units of Measurement and Analysis We have designed this guide specifically to support research on marriage and cohabitation determinants, although it also documents many survey features pertinent to other research topics. We describe here the key outcomes, determinants, and units of analysis the guide covers. Outcomes of interest: This guide assumes interest in understanding the forces leading people to decide to live together, to marry, and to break up, and the influences affecting the quality of their relationships. These outcomes reflect the federal government’s goals of promoting healthy marriage and strengthening relationships between unmarried couples. Thus, the key outcomes in the guide are the formation, dissolution, and quality of marriage and cohabitation. Formation and dissolution can be measured in a survey by ascertaining dates unions begin and end, or, more crudely, in terms of the union status at the time of the interview. Union quality can be measured in a myriad of ways, including self-reports of relationship satisfaction, self-reports on interaction behavior between partners, and—although it has not yet been attempted in a national survey—external observations of couple interaction. Determinants of interest: We assume that users of this guide are interested in testing hypotheses on the effects of key demographic, economic, socio-cultural and psychological factors on union formation, dissolution and quality. A wide variety of potential determinants are subsumed under each of these four headings. In our literature review (Fein et al. 2003), we identify some of the most important themes of concern to researchers. The literature review identifies ten major categories of influences. The first three influences cover key demographic processes: early and non-marital childbearing, transitions to marital parenthood, and non-marital cohabitation. These processes establish a framework for the remaining economic, socio-cultural, and psychological forces. Under the economic heading, we examine research on the effects of women and men’s economic fortunes. Important socio-cultural influences are the changing social significance of marriage and cohabitation, and the varying roles expected of men and women within relationships. Finally, three influences addressed in psychological research include couple interaction processes and the connections between these processes and personal characteristics and wider environmental contexts. For each of the nine surveys, we assess the degree to which they provide measures for important demographic, economic, socio-cultural, and psychological determinants. We also look across the surveys to ascertain possibilities for addressing questions arising in connection with each of the ten major categories of influences. Relevant units of measurement and analysis: Researchers in this field typically use the individual adult as the unit of analysis—to investigate, for example, influences on why some individuals get married or divorced. The union itself, however, may also be the unit of analysis, as in studies of marital quality. And some past studies have used community- or state-level aggregate measures (for 2 Chapter 1: Introduction Abt Associates Inc. example, marriage rates) as their dependent variables. This guide focuses on outcomes measured at or below the family level only. Explanatory variables can likewise be measured at a variety of levels. They include characteristics of the individual (for example, age), of the union (for example, date of formation), of the individual’s or union’s household (for example, total income relative to the federal poverty level), and of the individual’s or union’s community (for example, level of welfare benefits). Researchers will frequently want to combine information from multiple levels in the same analysis. For example, interest in the effects of community and other contextual influences on individuals often lead analysts to include descriptors of higher-level units in an analysis of a lower-level unit. Researchers may also do the reverse: include characteristics of the individual members of a union as explanatory variables when the dependent variable is measured at the level of the union, for example, husband’s employment, wife’s employment. Community-level data appears in standard data sets in two ways. Some surveys include community identifiers, enabling researchers to link information from other sources with more or less ease and specificity. (The Survey of Income and Program Participation, for example, identifies only state or, in the case of small samples, state groupings.) Other surveys explicitly include some communitylevel data. Several surveys delete community identifiers from the public use files, but make them available to researchers on special request. Desirable Attributes of Data Sets In addition to measuring key outcomes and influences for the correct units of analysis, there are a number of features of samples and survey designs needed to support strong analysis of union determinants. Our literature review identified a number of key attributes. Here, we recapitulate recommendations from the literature review and note the associated data requirements. 1. Study Influences within Disadvantaged Populations. The vast majority of analysis to date has paid little attention to interactions between influences and indicators of socioeconomic disadvantage other than race/ethnicity. Results thus are uninformative about the likely responsiveness to exogenous (policy-driven) changes in demographic, economic, socio-cultural, and psychological influences of populations of concern to policy. There is a compelling need for more comparative research on populations distinguished by characteristics such as family background, education, poverty, and neighborhood environment. Data sets thus should contain sufficient samples of disadvantaged populations, variously defined, to support analyses of determinants for these populations. 2. Include Both Men and Women in the Analysis. Better data collection on men has enhanced our ability to determine the effects of both (potential) partners’ attributes on union formation, stability and quality. Researchers are exploring the role of differences in characteristics between partners, as well as the degree to which the effects of one partner’s characteristics depend on the other partner’s characteristics. Analyses of union formation require identification of relevant potential partners, whereas for analyses of union stability and quality the actual identities of both partners are known. More research on the social networks defining marriage and cohabitation “markets” and data on Abt Associates Inc. Chapter 1: Introduction 3 both partners in more casual dating relationships are steps towards addressing this need. Surveys collecting data on both women and men, and ideally both partners in unions, are most useful. 3. Study Varied Types of Unions. Although the traditional marriage research paradigm stressed first marriages, the rise in divorce and remarriage has brought increased interest in step-family arrangements. The determinants and dynamics of later unions— cohabitations as well as remarriages—are likely to differ from those of first unions due to different kinds of relationships with children from prior relationships compared with biological children, and because experiences in prior unions are likely to affect behaviors in current ones. Compared with marriage, the forms of cohabitation may be more varied. There are likely to be different modal forms of cohabitation—differentiated in terms of frequency, exclusivity, long-term functioning, and other attributes. Particularly among disadvantaged groups, informal living arrangements are more fluid and may be perceived differently by each partner. Research on the determinants and consequences of cohabitation needs to identify and study the determinants of varied types of living arrangements. Surveys obtaining more detailed measures of union outcomes thus are most desirable. 4. Pay More Attention to Relationship Onset. Researchers have tended to study marriage and cohabitation as discrete states. A more satisfactory understanding of transitions to and from unions requires studying how partners meet and decide they are interested in each other, and the processes governing commitment and eventual decisions to cohabit and marry. Survey data covering dating relationships and inquiring about relationship development are especially helpful. 5. Refine Measures of Union Quality. Relationship quality is a complex construct, and yet it lies at the center of current policy interest in promoting healthy marriages. Researchers have used a myriad of definitions and measures—variously drawing on subjective assessments and objective characteristics of relationships—in studying influences on quality. Standardization on valid and reliable measures would help greatly in comparing results across studies and providing policy researchers with stronger measures for use in program evaluation. Surveys obtaining a wider range of behavioral reports and satisfaction assessments will be useful for developing improved quality measures. 6. Refine Measures of Certain Influences. A number of important influences require clearer definition and improved measurement. One example is the concept of the “importance” or “significance” of marriage and cohabitation. Indicators variously include general values, values applying to specific aspects of unions, benefits and obligations people believe derive from unions, and expectations they apply to their own relationships. Another concept needing refinement is “economic status,” which can be measured in material or social terms, as current or projected future quantities, and in absolute or relative terms (for example, in relation to consumption aspirations or partner’s status). In general, surveys that collect greater detail on one or more categories of influence will be most useful. 7. Analyze Effects over the Life Course. Researchers increasingly are exploiting panel data to examine the effects of factors measured at one point in time on later outcomes. 4 Chapter 1: Introduction Abt Associates Inc. These analyses often do not specify the ages at which outcomes occur, and the data often are weighted towards the early adult years. It thus is difficult to determine whether a measured response to a particular influence is a temporary effect (for example, postponed marriage) or a permanent response (for example, non-marriage). Improved analysis requires following sample members over a longer portion of the life course1 and testing interactions between age and (appropriately lagged) measures of influences. Surveys that collect retrospective data from older respondents and those that follow individuals and couples to older ages over time are needed for life course analyses. 8. Study Recent Experience and Compare Influences across Cohorts. Recent decades have witnessed dramatic changes in attitudes and behaviors regarding marriage and cohabitation. The substantial potential for change in the effects of determinants indicates a need for ongoing analysis of virtually every topic. Due to lags in data collection and analysis, most of what we currently know about influences on marriage and cohabitation is based on survey data from the 1980s. Because successive cohorts experience the same phases of life in different social and economic environments, cross-cohort comparisons of effects are needed to absorb fully the implications of period change. Researchers have made valuable contributions through cohort comparisons using data from successive cross-sectional surveys, a form of analysis that ought to continue in the future. Surveys that collect panel data for successive cohorts are the most responsive to this need. Different surveys of successive cohorts also can support cross-cohort analyses if their items are highly comparable. An important requirement is that surveys be ongoing, with data available for recent periods. 9. Study Interactions between Influences. Piecemeal study of various demographic, economic, socio-cultural and psychological determinants has led inescapably to the conclusion that the influences on marriage and cohabitation are many and complex. The next generation of research must devote substantial attention to overarching conceptual frameworks that accommodate all major influences and interactions between influences. Of particular importance is the need to specify the linkages between factors external to couple relationships and the internal relationship processes that directly affect union formation, stability and quality. Surveys collecting information on a variety of determinants are needed for richer analyses. 10. Develop and Apply Stronger Methods for Establishing Causation. During the past decade, researchers have made progress in addressing the chief threats to causal inference: selection bias and reverse causation bias. More sophisticated techniques have not always provided definitive answers, but they certainly have revealed much about the existence and nature of biases plaguing simpler multivariate analyses. Continued work in this area is highly desirable. Stronger conceptual frameworks, richer measures, and careful sensitivity testing are needed in analyses relying on controls for observed heterogeneity. Where possible, fixed effects models and natural experiments should be used to measure and adjust for unobserved heterogeneity. Temporally ordering 1 Retrospective history information from older respondents may suffice for some analyses, whereas long-term prospective designs will be needed in the many situations where poor recall compromises data quality. In general, analyses of first unions require measurement through at least peoples’ early 40s, and analyses of subsequent experiences require observation through at least their 50s. Chapter 1: Introduction 5 Abt Associates Inc. independent and dependent measures is a good initial step to clarifying the direction of estimated effects: more convincing analysis requires use of theoretically grounded models and structural equation methods. Finally, planned experiments where program impacts are limited to a single influence of interest provide exceptional opportunities to ascertain effects of a wide range of determinants (for example, teen childbearing, employment and earnings, type of neighborhood) on marriage and cohabitation, even when programs did not explicitly seek to alter union outcomes. Absent data from experiments, national longitudinal surveys offer the best prospects for fixed effects, natural experiments, and temporal ordering, especially those collecting data on both influences and outcomes on a relatively frequent basis. Data Sets Included in this Guide ACF conceived that this guide would focus on a limited number of national-level surveys. We chose for inclusion nine surveys that we felt offered exceptionally good opportunities for research on marriage and cohabitation determinants. None of these nine surveys possesses all of the attributes we might desire. Approaching this ideal would require a new survey vehicle, the need for which was an important recommendation from a recent national workshop on federal data collection on couples (Federal Interagency Forum on Child and Family Statistics 2002). Each of the nine surveys in this guide nonetheless is a major, ongoing national survey that offers strong prospects for studying at least some marriage and cohabitation determinants. These surveys include: • Current Population Survey (CPS). The CPS provides an annual demographic survey of large cross-sectional samples and is representative not only nationally but also by state and some sub-state areas. It oversamples Hispanics, other nonwhites, and households with children. Only limited information is available on union outcomes. Nonetheless, because the CPS has been existence since 1940, it offers unparalleled opportunities for analyses of time trends, including effects of state policies. Fragile Families and Child Well-Being Survey. Fragile Families is unique in that it interviews both parents of newborn children, regardless of their marital and co-residential status. Three-quarters of the sampled couples were unmarried at baseline. Follow-up interviews allow researchers to trace the subsequent evolution of relationships. Although it is representative only of couples with an infant in cities with population over 200,000, Fragile Families offers rich detail on the experiences and attitudes of both parents for this sample. National Longitudinal Study of Youth (NLSY79 and NLSY97). Starting in 1979, NLSY79 interviewed a nationally representative cohort of youth annually for 15 years, and biennially thereafter. By now, this cohort is around 40 years old. NLSY97, a second cohort that is currently about 20 years old, is being interviewed annually. Blacks and Hispanics were oversampled in both cohorts. A great wealth of background information is available on these samples, including unusual data on vocational ability and childhood living arrangements, and (for the 1979 cohort) multiple assessments of gender role attitudes. The long time span supports prospective analyses of union formation and dissolution for the 1979 cohort. Chapter 1: Introduction Abt Associates Inc. • • 6 • National Survey of America’s Families (NSAF). Comprising repeated cross-sectional surveys that are both nationally representative and representative for 12 states and the District of Columbia, NSAF offers good descriptions of families’ economic well-being and child welfare. It does not collect marriage or cohabitation histories, or information on union quality. Its greatest comparative advantage is for analyzing impacts of welfare reform, through linkages with a companion database on state policy measures. National Survey of Family Growth (NSFG). The NSFG is now in its sixth cycle, for the first time interviewing men as well as women. Cross-sectional in nature, the NSFG collects detailed information on marriage, cohabitation, and sexual histories, including partners’ demographic characteristics and respondents’ attitudes on men’s and women’s roles. Among its strengths are its oversampling of blacks and Hispanics, and its ability to analyze changes over its 30-year history. National Survey of Families and Households (NSFH). This longitudinal survey is probably the strongest single resource for analyzing union outcomes. Commencing in 1987/88 and now in its third wave of data collection, NSFH includes interviews with primary respondents, their spouses or cohabiting partners, their dependent and adult children, and their parents. NSFH oversamples blacks, Puerto Ricans, and Mexican Americans, and collects rich information on marriage and cohabitation histories, union quality, and attitudes towards marriage and cohabitation. Panel Study of Income Dynamics (PSID). The PSID has followed selected households since 1968, providing a detailed history of members’ experiences for over 30 years through annual interviews (biennial since 1997). Low-income households were oversampled. In contrast to the NLSY, the original PSID sample included individuals of all ages, and has grown by marriages and births among original sample members. It thus forms a microcosm of the United States. Although lacking attitudinal data, rich information is available on economic well-being, education, and time use. Union histories can be linked to background characteristics and the local economic and policy environments. Survey of Income and Program Participation (SIPP). The SIPP combines features of cross-sectional and longitudinal data. Its design is based on multiple independent panels, each followed for two to four years. Over the lifetime of each panel, detailed data are collected three times a year on household composition, income, employment, and participation in means-tested programs, while a set of topical modules accumulates a wealth of information on marital history, household relationships, work schedules, and other subjects. These data are thus excellent for prospective analysis of relatively shortterm processes. Survey of Program Dynamics (SPD). The SPD builds on the 1992 and 1993 panels of the SIPP, surveying households annually for a total of ten years. It draws on the wealth of background data collected by the SIPP, and also includes annual measures of union quality (for couples with children). The 10-year span supports prospective analysis of relatively long-term union outcomes. Severe sample attrition may, however, pose a problem for representativeness. • • • • • Abt Associates Inc. Chapter 1: Introduction 7 Resource limitations did not allow us to include a number of other surveys that also might be good choices for certain analyses. In particular, researchers may wish to explore analysis opportunities associated with the following surveys: • American Community Survey. This ongoing survey, and the decennial census long form survey which it will replace, provide measures of current marital status and cohabitation status and basic socio-economic characteristics. The data will support crosssectional time series analysis and estimation for detailed subpopulations and local areas.2 Decennial Census. Decennial census data provide another source for cross-sectional analyses of marriage and cohabitation status requiring estimates for detailed subpopulations and local areas. In 2000, dropping the direct question on marital status for all people left analysts able to identify only married couples where one is the householder.3 Early Childhood Longitudinal Study. Prospective surveys of recent kindergarten and birth cohorts will provide substantia l information on children’s family contexts, including detailed information on their parents’ unions. The emphasis on child development and well-being makes these surveys good choices for studies of how union outcomes for parents affect children. 4 General Social Survey. An annual cross-sectional survey on a nationally representative sample of about 1,500 adults conducted nearly every year since 1972, the GSS offers opportunities to examine a time series in a number of important values and attitudes bearing on marriage and related aspects of family life.5 Intergenerational Panel Study of Parents and Children. This survey, an eight-wave intergenerational survey (1962-1993) of white couples in the Detroit area who had a birth in July 1961, offers exceptionally rich measures on attitudes and other predictors, as well as details on union formation, dissolution and quality. 6 Marital Stability Over the Life Course Study. This ongoing national survey is designed to study the determinants and consequences of marriage attitudes and behaviors of about 2,000 married persons (not couples). The sample has been interviewed by telephone periodically since 1980. Interviews with adult offspring of the original survey sample in 1992 and 1997 provide unique opportunities to study inter-generational effects.7 National Education Longitudinal Surveys. Surveys under this program have followed education, career, and family outcomes of young people as they moved into their early adult years. Data collection for two of these surveys ended in the 1980s (National • • • • • • 2 3 4 5 6 7 For further information, see www.census.gov. For further information, see www.census.gov. For further information, see www.nces.ed.gov. For further information, see www.socio.com. For further information, see www.socio.com. For further information, see www.pop.psu/marinst. Chapter 1: Introduction Abt Associates Inc. 8 Longitudinal Study of the High School Class of 1972) or early 1990s (High School and Beyond), but the third survey is ongoing (National Education Longitudinal Study of 1988). They offer excellent opportunities to study the effects of education and early career development on union experiences during the young adult years.8 • National Household Education Surveys. This ongoing program has surveyed large (45,000 to 60,000), nationally representative cross-sectional samples of households about every two years since 1991. Their strength is the ability to relate household members’ marital status and in selected waves, cohabitation status, to education and child care outcomes for children and to the educational and career-related outcomes of adults.9 Welfare, Children, and Families: A Three City Survey. This panel study, begun in 1999, involves approximately 2,400 households with children in low-income neighborhoods in Boston, Chicago, and San Antonio. Designed to assess experiences with welfare reform, this survey is collecting rich measures of children’s well-being and family dynamics. An additional feature is comparative ethnographic studies of a subsample of 215 families.10 • Contents of this Guide The body of this report is comprised of separate chapters on each of the nine selected surveys. Separate sections detail key aspects of each survey, including its design, general content, utility for research on marriage and cohabitation outcomes and determinants, and data availability. Accompanying exhibits provide additional detail on survey content. Where possible, these exhibits have been taken directly from survey documentation. In the concluding chapter, we provide discussion across the nine surveys of the degree and ways in which they can be used to address major gaps in knowledge of marriage and cohabitation determinants. Sections correspond to the ten broad themes we investigated for the literature review component of this project. 8 9 10 For further information, see www.nces.ed.gov. For further information, see www.nces.ed.gov. For further information, see www.jhu.edu/welfare. Chapter 1: Introduction 9 Abt Associates Inc. Chapter Two: The Current Population Survey The Current Population Survey (CPS), under the direction of the U.S. Census Bureau, is the primary data source for measuring unemployment and labor force participation for the United States as a whole and for the individual States. The March supplement to the CPS, officially called the Annual Demographic Supplement, collects detailed data on households’ financial circumstances and wellbeing. 11 Information on many other household and individual characteristics is collected with varying periodicity. The survey is ongoing. Survey Design The CPS is representative both nationally and by state, and also for substate areas in New York and California. A complex stratified random sample survey, the sample changes continuously as new addresses are brought into the survey and old addresses are retired. Each sampled household is interviewed for four months, then “rests” for eight months, then is interviewed for four more months, for a total of eight interviews over 16 months. It is important to note that the interview unit is the address, not the people living there. If a household moves away, the survey replaces the respondents with the new people living at that address. One person per household (the reference person) provides proxy information for everyone living at that address. Demographic data are collected for all household members, and labor force data for all members aged 15 and over. The sample for the March supplement includes not only all households active in March, but also inactive households that were interviewed in November or April headed by Hispanics or other nonwhites, and other households with children aged 18 or younger. The samples in these subgroups are thus substantially strengthened. Appropriate sample weights are available. Sample sizes: The sample size in each month comprises about 48,000 housing units. Methods of data collection: All interviews are computer assisted, either CAPI or CATI. The first and fifth interviews are conducted in person, and 85 percent of the remainder by telephone. Response rates: Response to the CPS varies by month-in-sample. In September 2002, nonresponse rates were about 8 percent for households in their first and fifth months (i.e. the first contact and the first contact after the four month break), around 6 percent for months two through four, and 6 to 7 percent for months six through eight. Survey Content A household roster is established or updated at each interview. Information collected for each household member includes age, sex, marital status, education, race, and relationship to the reference 11 Other demographic data are collected monthly on the household roster. Chapter 2: The Current Population Survey 11 Abt Associates Inc. person (such as spouse, child, grandchild, parent, brother/sister, other relative, foster child, roommate, unrelated, unmarried partner). In addition, the spouse and one parent of each household member are identified, if these are household members. The basic monthly survey collects detailed information on labor force participation, including ownership of a business or farm, work for pay or profit, unpaid work, layoffs, job search activitie s, temporary absence from work, number of jobs, hours worked, reason for part-time work, reason for neither working nor seeking work, occupation, sector, type of business, usual work activities, earnings, overtime pay, tips, and commissions. Citizenship and nativity information are also collected. The March supplement covers the following topics: total family income, individuals’ work experience, earned income, unearned income, health insurance, employers’ pension plans, non-cash benefits, welfare program services, and migration. Every five years the CPS administers the Fertility and Marital History supplement. This instrument collects information on the date of birth, sex, and current living arrangements of all children born to the respondent, and the dates on which each marriage began and ended. The most recent administration was March 2001, with the next planned for 2006. In recent years other supplements have been conducted on a broad range of primarily economic topics. Supplements of greatest potential relevance to research on union outcomes include those on race and ethnicity, job tenure and occupational mobility, child support, and work schedule and homebased work. Suitability for Research Measures of outcomes: Marital status of each household member (married, spouse present, married spouse absent, widowed, divorced and never married) is identified on the household roster and updated at each contact. Starting in 1995, the CPS began collecting information directly on cohabitation status by including “unmarried partner” as a relationship type on the household roster. Changes in union status can be observed over the 16 months of coverage. Specific dates of marriage transitions are also reported in the quinquennial Marriage and Fertility Supplement prior to 2001. There are no measures of union quality available in the CPS. Measures of demographic influences: The Marriage and Fertility Supplement provides information about parent’s age at child’s birth that is relevant to studying the influence of teen childbearing on union formation and dissolution. The study of non-marital births is limited by the lack information on cohabitation and marriage history, exact parentage of each child in a household, and parents’ relevant values and attitudes. Similarly, the lack of union satisfaction and interaction measures poses a barrier to using these data for studying transitions to parenthood, despite the availability of fertility histories. Basic information is available for studying effects of cohabitation on marriage, but not in the context of relationship quality, time spent together, values, or expectations. The current status measures available from the household roster provide no information about union tenure, which has been found to be an important covariate for understanding the influence of prior cohabitation on marriage outcomes. 12 Chapter 2: The Current Population Survey Abt Associates Inc. Measures of economic influences: The CPS provides monthly data on job characteristics and household composition for both men and women. The March Supple ment provides further information on work experience, and the May 1997 Work Schedule Supplement includes details on multiple job holding and hours of work. These features suggest the potential of these data for exploring effects of both men’s and women’s employment on union behaviors. Measures of socio-cultural influences: The CPS provides limited opportunities for socio-cultural research on union outcomes. It does not collect data on respondents’ views on marriage and cohabitation or men’s and women’s roles, so that the influence of these factors on union behaviors cannot be examined directly. However, the measures of male and female employment and earnings available in the CPS, as well as fertility and household structure data, have the potential to describe men and women’s breadwinning and household labor behaviors and link these with union behaviors. This may offer indirect evidence of the influence of gender roles on union outcomes. Measures of psychological influences: The CPS is an inappropriate data set for direct examination of psychological factors in union outcomes, as it does not collect data on relevant measures such as behaviors within relationships, feelings and emotions, or relationship satisfaction. Special strengths: The CPS has been used primarily for nationally representative descriptive analyses, and is a central source of information on demographic trends and patterns related to unions. It supports aggregate time-series/cross-section analysis by state, going back for several decades, allowing analysis of state-level policy measures. The CPS has excellent monthly data on employment and earnings. Combining the occasional supplements with the Annual Demographic Supplement allows the analyst to explore the influences of particular economic and employment issues such as work schedule, job tenure, and occupational mobility. The rich array of measures around individual and household economic well-being in CPS could be capitalized upon to investigate differing approaches to defining and describing disadvantaged populations and their union patterns. Limitations: The CPS is limited in its usefulness for describing transitions in union outcomes. Although longitudinal in theory, the survey is essentially cross-sectional for some purposes because addresses, not households, are the sample units. If a family moves away, its members are lost to the survey, and the new occupants of the residence are interviewed in their stead. Although limited marital histories were collected in prior years, the June 2000 CPS did not collect marital histories, and there has been a strong call for the return of this measure so that researchers can measure trends in family formation (Federal Interagency Forum on Child and Family Statistics, 2002). The CPS does not collect a cohabitation history or any measures of relationship quality. Beyond economic and labor force measures, and information about other household members, the CPS is relatively weak in its availability of other potential explanatory variables (such as attitudinal measures). Using the Data The website for the CPS is found at http://www.bls.census.gov/cps. The home page provides links to descriptions of the survey, user guides, reports, and downloadable data files. Abt Associates Inc. Chapter 2: The Current Population Survey 13 Documentation: A key resource for potential users, available on the CPS website, is: CPS Technical Paper 63 RV: Design and Methodology. US Department of Labor, Bureau of Labor Statistics and US Department of Commerce Economics and Statistics Administration, US Census Bureau, March 2002. Data availability: Public use microdata files containing the basic labor force and demographic data, along with supplemental data, are available for sale from the Customer Services Branch of the Census Bureau. They can be purchased in either tape or CD-ROM format. In addition, data can be downloaded from the CPS website via FERRET. Reports and articles: An extensive list of documents, with links, is available at: http://www.bls.census.gov/cps/pubsmain.htm. A few examples of studies using the CPS to analyze union formation and dissolution are: Bitler, M., Gelbach, J. B., & Hoynes, H. W. (2002). The impact of welfare reform on living arrangements. NBER Working Paper Series, Working Paper No. 8784. Casper, L. M. & Cohen, P. (2000). How does POSSLQ measure up? Historical estimates of cohabitation. Demography, 37(2), 237-245. Lichter, D. T, McLaughlin, D. K., & Ribar, D. C. (2002). Economic restructuring and the retreat from marriage. Social Science Research, 31(2), 230-256. Mare, R. D. & Winship, C. (1991). Socioeconomic change and the decline of marriage for blacks and whites. In C. Jencks and P. E. Peterson (Eds.), The Urban Underclass (pp. 175202). Washington, D.C: The Brookings Institution. Data Elements The content of the March supplement is shown in Exhibit 2.1. The content of the Fertility and Marital History Supplement is shown in Exhibit 2.2. 14 Chapter 2: The Current Population Survey Abt Associates Inc. Exhibit 2.1 CPS: Content of Annual Demographic Supplement (March Supplement 2001) Family Income Total combined income for past twelve months. Work Experience Employment status, reasons for not working, number of employers, hours worked per week, number of weeks worked part-time (35 hours or less), reasons for part-time work, identification of longest job (name of employer, industry, occupation, duties, number of employees). Earned Income Unemployment and Workers’ Compensation Social Security Social Security for Children Supplemental Security Income (SSI) Supplemental Security Income for Children Public Assistance Veterans Payments Survivor Benefits Disability Income Retirement and Pensions Dividends Property Income Education Assistance Child Support and Alimony Regular Financial Assistance Other Money Income Health Insurance Employer’s Pension Plan School Lunches Public Housing Food Stamps Energy Assistance New Welfare Reform Transportation assistance (i.e. gas vouchers, bus passes, or help repairing vehicle), child care services or assistance, attendance of GED classes or basic skills training, attendance of a job search program, specific job skill training, participation in a work experience program. Migration Residence on March 1, 2000, main reason for moving. Abt Associates Inc. Chapter 2: The Current Population Survey 15 Exhibit 2.2 CPS: Content of the Fertility and Marital History Supplement (1995) Fertility Number of live births. For each child: residence, birth date, age, gender, R’s age at birth of child. Marital History Number of marriages. For each marriage: divorce, widowhood, date of marriage, R’s age at time of marriage, date cohabitation ended, date marriage ended. 16 Chapter 2: The Current Population Survey Abt Associates Inc. Chapter Three: The Fragile Families and Child Wellbeing Study Fragile Families is an ongoing survey of a cohort of families with a new child. Three-quarters of the sample comprises families in which the child’s parents were not married to each other at the time of the birth (“fragile families”); the remainder of the sample comprises families headed by a married couple. The study includes a baseline survey conducted in 1998-2000 shortly after the focal child’s birth, and follow-up interviews at 12, 30, and 48 months. Both parents are interviewed each time. The primary purpose of this study is to provide information on nonmarital childbearing, welfare reform, and the role of fathers in fragile families. The survey is sponsored by the Center for Research on Child Wellbeing (Princeton University) and the Social Indicators Research Center (Columbia University School of Social Work). Several governmental agencies and private foundations fund the project. These include the National Institute of Child Health and Human Development, the Office of Population Research, the National Science Foundation and the U.S. Department of Health and Human Services (ASPE and ACF). Survey Design The survey is being conducted in 20 sites. Sixteen of the 20 sites were selected as a stratified random sample of cities with population over 200,000. The stratifying variables used were measures of welfare generosity, the child support system, and the local labor market. Data from these sites are nationally representative of births in cities with population over 200,000, and are referred to as the “nationally representative sample.” The remaining four sites were chosen because they were of special interest to sponsoring foundations : Newark, NJ, Oakland, CA, Detroit, MI, and San Jose, CA. Inclusion of data from these sites produces the “full sample.” Families are identified for inclusion in the survey at the hospital at the time of their infant’s birth. Hospitals were randomly chosen within cities, and births within hospitals. Sample sizes: The sample size is 4,898 families, including 3,712 unmarried couples and 1,186 married couples. Methods of data collection: The mother’s baseline interview was typically administered in person, in the hospital. If the mother was discharged from the hospital before she could be interviewed, she was surveyed by phone. Fathers were interviewed either in person in the hospital, in person at home, or by telephone. Hospital medical records were collected for both mothers and infants. Response rates: Baseline response rates for eligible fathers and mothers varied by marital status, but were quite similar for the full and nationally representative samples. Abt Associates Inc. Chapter 3: The Fragile Families and Child Wellbeing Study 17 Full Sample Mothers Unmarried Married Fathers Unmarried Married 87% 82% 75% 89% Nationally Representative Sample 87% 82% 76% 89% Response rates for the 12-month followup have not yet been released. Survey Content Mothers and fathers are interviewed individually at each contact. The instruments used are similar but not identical. For both mothers and fathers, the baseline instruments include sections on child health and development, father-mother relationships, fatherhood, marriage attitudes, relationships with extended kin, environmental factors and government programs, health and health behavior, demographic characteristics, education and employment, and income. The fathers’ instrument includes an additional section on work activities. The 12-month follow-up instruments include sections on family characteristics, child well-being and mothering (fathering), father-(mother-)child relationship, mother’s relationship with father (father’s relationship with mother), current partner, demographics, mother’s (father’s) family background and support, environment and programs, health and health behavior, education and employment, and income. The variations in parentheses refer to the fathers’ instrument, in contrast to the mothers’ instrument. That is, each parent is asked about child well-being and their own parenting, about the relationship of the other parent to the child, about their relationship with the other parent, and about their own family background and support. The three-year follow-up instruments include the same sections, plus an additional section on word associations (for example, “In what way are a dog and a lion alike?”) designed to measure intelligence. In addition to data on child health and development colle cted from the parents during each of the follow-up interviews, in-home assessments of child well-being will be carried out at 30 and 48 months. Child well-being measures will overlap with those used in other national studies, including the Infant Health and Development Program, the Early Head Start Evaluation, the Teenage Parent Demonstration Project, and the Early Childhood Longitudinal Study - Birth Cohort Study. Suitability for Research Outcome measures: The survey collects data on union status of both mother and father at each interview. Changes in union status subsequent to the birth of the focal child can readily be traced. No historical data are available , however. Rich information is available on union quality, including the couple’s relationship with each other and each member’s relationship with the focal child. Measures of demographic influences: Fragile Families is an obvious choice for studying the effects on marriage of teen and non-marital childbearing, transitions to parenthood, and cohabitation, 18 Chapter 3: The Fragile Families and Child Wellbeing Study Abt Associates Inc. despite the lack of marital and cohabitation histories. Rich information is available about the union that produced the focal child, including relationship between the partners, relationship between each partner and the child, and positive and negative interactions. The cautionary note remains the representativeness of the subsamples relevant for studying each of these questions. Measures of economic influences: In addition to standard measures of employment, Fragile Families includes some unusual detail with regard to employment: reservation wage, off-the-books work, illegal activities for money and rich measures about welfare. Because of the sample structure, it is not appropriate for investigating union formation, but is a good source for effects of men’s and women’s employment and employability on transitions from cohabitation to marriage, and union dissolution. The influence of the marriage market on these same behaviors could potentially be addressed; aggregate data about the marriage market from another source could be added to the individual-level data from Fragile Families, since the city of residence is known. Measures of socio-cultural influences: Fragile Families offers an opportunity to examine the influence of key attitudes on union behaviors, for its unique sample members. Attitudinal data is collected from both men and women on gender roles (Exhibit 3.7), and the meaning of marriage and cohabitation (Exhibit 3.8). Attitudes on fatherhood are collected as well (Exhibit 3.9). The data on attitudes are collected at baseline. This permits analysis of differences between individuals at the time of their child’s birth, and the information can be combined with the longitudinal data to examine the influence of pre-existing attitudes on later union behaviors. Additionally, given the collection of attitudinal measures from both partners, analyses can consider the extent of discordance between partners’ attitudes and its influence on their union formation (that is, cohabiting couples subsequently marrying, and non-cohabiting couples starting to live together), as well as union dissolution and quality. The extent to which attitudes mediate other influences on unions, such as employment opportunities or parenting, can also be examined due to the availability of an array of information on these other relevant behaviors. Measures of psychological influences: Fragile Families has good data on behaviors and satisfaction, from both partners, and includes measures of some important contextual influences. It includes measures of depression and anxiety for both partners. Special strengths: Fragile Families is a particularly strong data set for examining union behaviors, as it was designed with this research area as one of its primary concerns. This study is unique in that it interviews both parents of newborn children, regardless of their marital and co-residential status. Furthermore, the follow-up interviews allow researchers to trace the evolution of relationships, even for couples that are not living together. Fragile Families collects detailed attitudinal information directly from both fathers and mothers. Together these design elements allow couple -based analyses, incorporating measures collected directly from each partner. This is a great strength as compared to designs that collect proxy information about a partner from the respondent. Additionally, the design of Fragile Families still permits more traditional individual level analysis of men or women’s experiences. An additional strength of Fragile Families is its focus on unmarried couples with babies. This makes it especially suitable for analysis of union behaviors among this disadvantaged subpopulation, which is of great policy interest. This focus on a specific subsection of the population is further strengthened by its very well-specified collection of measures about potential influences on union formation for disadvantaged individuals, such as welfare issues, child support, and support from Abt Associates Inc. Chapter 3: The Fragile Families and Child Wellbeing Study 19 extended families. There is greater breadth and depth of concepts and measures relevant to disadvantaged families than would be available in a survey designed to measure experiences in a more general population (such as the NSFG). At the same time, the inclusion of couples married at the time of their child’s birth offers a statistical control group that usefully expands the range of questions and comparisons the data can address. Beyond the couple -level analysis noted above, the Fragile Families data set lends itself to other analytical techniques which may offer needed leverage for better addressing issues of unobserved heterogeneity or contextual effects that plague most causal analyses. First, the collection of multiple observations from each respondent permits the use of statistical approaches such as linear growth models designed to describe individual trajectories, which would be particularly appropriate for examining union formation and dissolution experiences. Second, since the sample is limited to births in 20 different cities, there is adequate variation and sample size for fixed-effects models that may be especially useful for considering the effect of contextual influences, such as community norms and attitudes, otherwise difficult to directly control. (Although the public use file does not identify sites, this information can be obtained by special request.) Limitations: Many of the strengths of Fragile Families are, at the same time, its limitations. It does not purport to survey the general population; it is a special-purpose study including only couples with a new infant in large urban areas. Hence it cannot be used to answer general questions about union formation and dissolution. More specifically, the data are not appropriate for examining union formation behaviors of individuals without children or of individuals residing in smaller towns or non-urban areas. As more research is made available utilizing this unique data set, it will be extremely important that these limitations be kept in mind when generalizing from their results. Furthermore, even within the context of the central relationship, between the mother and father of the focal child, the information collected is incomplete. The status and quality of the relationship is reported at each contact, but the previous history of the relationship is not addressed. Thus, there is no information on how long the couple has been together, how they met, or whether they cohabited before marriage. For couples that are not married or cohabiting at the time of the birth of the child, the survey does not determine whether they were previously married or cohabiting. Also, because the household roster does not distinguish between the respondent’s biological and stepchildren, it is not possible to identify joint offspring of the couple other than the focal child. Sources of household income are noted for the respondent only. The relatively low response rate from unmarried fathers noted above (75 percent) is also a potential source of concern. Response rates were especially low for non-cohabiting fathers. In only 60 percent of non-cohabiting couples for which a baseline mother interview was completed did the father also respond. If the father did not visit the mother in the hospital, the response rate was even lower: 35 percent. Using the Data The survey website is http://crcw.princeton.edu/fragilefamilies/index.htm. 20 Chapter 3: The Fragile Families and Child Wellbeing Study Abt Associates Inc. Documentation: The primary source for users is: Reichman et al. (2001). “The Fragile Families and Child Wellbeing Study: Sample and Design,” Children and Youth Services Review, 23(4/5), 303-326. This report is also available through a link on the Fragile Families website. Documentation of the 12 month and 30 month followup is not yet available. Data availability: Upon registration at the Office of Population Research (OPR) Data Archive, users can download the baseline data, as STATA, SPSS, or ASCII files. Release of the 12-month followup data is planned for Spring 2003, and the 30-month follow-up data are expected in Fall 2004. Fiveyear follow-up data are expected in Winter 2006. The downloadable files do not include city or hospital identifiers. This information is available to researchers by special request. Reports and articles: Two key reports using the baseline and 12 month follow-up data, respectively, are: Carlson, M., McLanahan, S., & England, P. (2002). Union formation and dissolution in fragile families. Center for Research on Child Wellbeing, Princeton University Working Paper No. 01-06-FF. Mincy, R. & Dupree, A. T. (2001). Can the next step in welfare reform achieve PRWORA’s fourth goal?: Family formation in fragile families. Children and Youth Services Review, 23(6/7), 577-601. Data Elements The tables that follow show the availability of specific information, organized as follows: I. Baseline Survey A. Mother’s Instrument (Exhibit 3.1) B. Father’s Instrument (Exhibit 3.2) Twelve-Month Followup A. Mother’s Instrument (Exhibit 3.3) B. Father’s Instrument (Exhibit 3.4) Three-Year Followup A. Mother’s Instrument (Exhibit 3.5) B. Father’s Instrument (Exhibit 3.6) II. III. Abt Associates Inc. Chapter 3: The Fragile Families and Child Wellbeing Study 21 IV. Attitudinal Questions A. Gender roles (Exhibit 3.7) B. General attitudes toward marriage, keys to a successful marriage (Exhibit 3.8) C. Attitudes towards fatherhood (Exhibit 3.9) 22 Chapter 3: The Fragile Families and Child Wellbeing Study Abt Associates Inc. Exhibit 3.1 Fragile Families: Content of Baseline Mothers’ Instrument Section A: Child Health and Development Child’s date of birth, gender, relationship to FATHERa (did father visit the hospital, duration of relationship before pregnancy, married/cohabiting), child’s expected living situation, presence of other biological children, birth weight, handling of birth costs, child’s resemblance.b Section B: Father-Mother Relationships Relationship with FATHER (before/during/after pregnancy, activities in last/past month), reasons for relationship termination, subjects of disagreement, frequency of certain behaviors, plans to cohabit/marry, views of marriage, support from FATHER. Section C: Fatherhood Opinion of important activities between father and child, opinion on rights and obligations of fathers. Section D: Marriage Attitudes General attitudes toward marriage, keys to a successful marriage, roles of men and women. Section E: Relationships with Extended Kin Household roster (name, relationship to R, gender, age, typical employment status), support from friends and family during pregnancy. Section F: Environmental Factors and Government Programs Current housing situation (owned/rented, public housing, safety), religion (affiliation and attendance), presence of welfare. Section G: Health and Health Behavior Opinion of own health, substance use/abuse. Section H: Demographic Characteristics Place of birth, race/ethnicity. Section I: Education and Employment Educational attainment, date last employed for two consecutive weeks, hours worked per week, earnings, FATHER’S educational attainment, FATHER’S race/ethnicity, FATHER’S age, FATHER’S employment, FATHER’S health, FATHER’S substance use/abuse, FATHER’S current residence. Section J: Income Presence of income from various sources, amount of income from various sources,b total household income for past twelve months, presence of a personal vehicle.b a FATHER refers to biological father of the focal child. b These questions were asked in “18 Cities” only. Abt Associates Inc. Chapter 3: The Fragile Families and Child Wellbeing Study 23 Exhibit 3.2 Fragile Families: Content of Baseline Fathers’ Instrument Section A: Child Health and Development Present at birth, R’s name on birth certificate, number of other biological children, child’s resemblance,b relationship to MOTHERa (before/during/after pregnancy, activities in last month, subjects of disagreement, frequency of certain behaviors), reasons for relationship termination. Section B: Father-Mother Relationships Cohabitation status, plans to marry/cohabit, activities in the past month, subjects of disagreement, frequency of certain behaviors, views of marriage, support of MOTHER during pregnancy, expected involvement in child raising. Section C: Fatherhood Attitudes toward fatherhood, opinion of important activities between father and child, opinion on rights and obligations of fathers. Section D: Marriage Attitudes General attitudes toward marriage, keys to a successful marriage, roles of men and women. Section E: Relationships with Extended Kin Household roster (name, relationship to R, gender, age, typical employment status), family structure, support from friends and family during MOTHER’S pregnancy. Section F: Environmental Factors and Government Programs Current housing situation (owned/rented, public housing, safety), religion (affiliation and attendance), welfare and child support policies in R’s city. Section G: Health and Health Behavior Opinion of own health, substance use/abuse, mental health. Section H: Demographic characteristics Place of birth, race/ethnicity. Section I: Education and Employment Educational attainment, job skill training, job search (how long, reservation wage), date last employed for two consecutive weeks, hours per week, job activities, self-employed,b earnings, employer-provided insurance, income from illegal activities. Section K: Income Presence of income from various sources, amount of income from various sources,b total household income for past twelve months, presence of a personal vehicle.b a MOTHER refers to mother of the focal child. b These questions were asked in “18 Cities” only. 24 Chapter 3: The Fragile Families and Child Wellbeing Study Abt Associates Inc. Exhibit 3.3 Fragile Families: Content of Twelve-Month Follow-Up Mothers’ Instrument Section A: Family Characteristics Name of child’s father, child’s date of birth, child’s living situation, relationship with FATHER, date of marriage to FATHER, start date of cohabitation with FATHER, date romantic relationship with FATHER ended, reason for relationship termination, current relationship with FATHER. Section B: Child Wellbeing and Mothering R’s view of child’s health and development, breastfeeding, separations between R and child (dates, child care during separation, reasons for separation), characteris tics of child, interaction with child, views on parenthood, childcare arrangements, money/voucher/scholarship for childcare. Section C: Father-Child Relationship Interaction between FATHER and child (frequency, activities), legal arrangements with FATHER (paternity, child support, visitation), informal child support agreements, number of children with someone other than FATHER, child support from other fathers, FATHER’S support of children other than focal child, FATHER’S current marital status, FATHER’S current residence, is FATHER currently in jail (offense), FATHER’S health, FATHER’S substance abuse. Section D: Mother’s Relationship with FATHER (for mothers who are or were in a relationship) Opinion of FATHER as a father, R’s trust of FATHER, rapport with FATHER, frequency of certain behaviors, abuse. Section E: Current Partner Gender, marital/cohabitation status, partner’s interaction with child (activities, discipline), R’s trust of partner, frequency of certain behaviors, abuse. Section F: Demographics Size of household, household roster (name, gender, age, relationship to R, employment status), number of children living outside household. Section G: Mother’s Family Background and Support Birthplace of father/mother, citizenship, educational attainment of father/mother, financial assistance from friends/family, extent of R’s support group, family structure, current relationship with parents, FATHER’S relationship with parents, R’s relationship with FATHER’S parents, how often does child see R’s parents, how often does child see FATHER’S parents. Section H: Environment and Programs Current housing situation (house/apartment/condo, rent/mortgage amount, public housing), assistance from government agencies (i.e. Head Start, WIC, AFDC/TANF, Food Stamps, Unemployment insurance, Workers’ compensation, SSI, EITC, etc.), problems due to lack of funds (i.e. need to seek free food, eviction, borrow money from friends, forgone medical treatment, etc.). Section J: Health and Health Behavior Opinion of health status, covered by federal, state, or private medical care agency, does this agency cover the children, smoking incidence, alcohol use/abuse, drug use/abuse, depression (presence, duration, intensity, side affects), anxiety (presence, duration, intensity, side effects). Section K: Education and Employment Current educational activity, educational attainment, job search, reservation wage, date last worked for two consecutive weeks, earnings, work hours, information on second job, information on “off-the-books” work, number of regular jobs in last three years, own business, illegal activities for money. Section L: Income Total household income, money given or loaned to friends, bank account (primary holder), credit card (primary holder), who handles financial responsibilities, car/truck/van (amount owed, resale value), what does R enjoy doing with child the most. Abt Associates Inc. Chapter 3: The Fragile Families and Child Wellbeing Study 25 Exhibit 3.4 Fragile Families: Content of Twelve-Month Follow-Up Fathers’ Instrument Section A: Family Characteristics Child’s date of birth, R’s date of birth, Child’s current living situation, Father’s visitation with non-residential children, relationship with MOTHER,a date of MOTHER’S death, date of marriage to MOTHER, start date of cohabitation with MOTHER, date romantic relationship with MOTHER ended, reason for relationship termination, current relationship with MOTHER, number of children with MOTHER. Section B: Child Wellbeing and Fathering R’s view of child’s health and development, age child was weaned, separations between R and child (dates, child care during separation, reasons for separation), characteristics of child, interaction with child (activities, discipline), views on parenthood, childcare arrangements, money/voucher/scholarship for childcare. Section C: Mother-Child Relationship Interaction between MOTHER and child (frequency, activities, discipline), legal arrangements with MOTHER (paternity, child support, visitation), informal child support agreements, MOTHER’S financial support, number of non-residential children, child support for other children. Section D: Father’s Relationship with MOTHER (for fathers who are or were in a relationship) Opinion of MOTHER as a mother, R’s trust with MOTHER, rapport with MOTHER, frequency of certain behaviors. Section E: Current Partner Gender, marital/cohabitation, partner’s interaction with child (activities, discipline). Section F: Demographics Size of household, household roster (name, gender, age, relationship to R, employment status), number of children living outside household. Section S: Word Association Section G: Father’s Family Background and Support Birthplace of father/mother, citizenship, educational attainment of father/mother, race/ethnicity, attendance of religious services, financial assistance from friends/family, extent of R’s support group, family structure, current relationship with father/mother, relationship between R’s parents and MOTHER, R’s relationship with MOTHER’S parents, how often does child see R’s parents, how often does child see MOTHER’S parents. Section H: Environment and Programs Current housing situation (house/apartment/condo, rent/mortgage amount, public housing), assistance received from government agencies (i.e. Head Start, WIC, AFDC/TANF, Food Stamps, Unemployment Insurance, Workers’ Compensation, SSI, EITC, etc.), problems due to lack of funds (i.e. need to seek free food, eviction, borrow money from friends, forgone medical treatment, etc.), legal charges pending, incarceration history. Section J: Health and Health Behavior Opinion of health status, covered by federal, state, or private medical care agency, does this agency cover the children, smoking incidence, alcohol use/abuse, drug use/abuse, depression (presence, duration, intensity, side affects), anxiety (presence, duration, intensity, side affects). Section K: Education and Employment Educational attainment, current educational activity, work history (for the year following 20th birthday), job search, reservation wage, work leave since birth of child, current/most recent job, work hours, information on second job, information on “off-the-books” work, number of regular jobs in last 3 years, own business, illegal activities for money. Section L: Income Total household income, money given/loaned to friends, bank account (primary holder), credit card (primary holder), who handles financial responsibilities, car/truck/van (amount owed, resale value), what does R enjoy doing with child the most. a MOTHER refers to biological mother of the focal child. 26 Chapter 3: The Fragile Families and Child Wellbeing Study Abt Associates Inc. Exhibit 3.5 Fragile Families: Content of Three-Year Follow-Up Mothers’ Instrument Section A: Family Characteristics Child’s current living situation, Mother’s visitation with non-residential children, relationship with FATHERa,a date of FATHER’s death, date of marriage to FATHER, start date of cohabitation with FATHER, date of divorce from FATHER, date romantic relationship with FATHER ended, reason for relationship termination, current relationship with FATHER, subsequent children since birth of focal child, number of children with FATHER, number of children with different father(s), number of romantic relationships before FATHER, pregnancies in these relationships and outcomes. Section B: Child Wellbeing and Mothering R’s view of child’s health and development, separations between R and child (dates, child care during separation, reasons for separation), interaction with child (activities, discipline), views on parenthood, childcare arrangements, money/voucher/scholarship for childcare. Section C: Father-Child Relationship Opinion of FATHER’s relationship with child, interaction between FATHER and child (frequency, activities, discipline), legal arrangements with FATHER (paternity, child support, visitation), informal child support agreements, FATHER’S support of children other than focal child, FATHER’S current marital status, FATHER’S current residence, is FATHER currently in jail (offense), FATHER’S health, FATHER’S substance abuse. Section D: Mother’s Relationship with FATHER (for mothers who are or were in a relationship) Opinion of FATHER as a father, R’s trust of FATHER, rapport with FATHER, frequency of certain behaviors, abuse, reasons for disagreements. Section E: Current Partner Gender, marital status, date of marriage, date cohabitation began, length of romantic involvement, date of birth, race/ethnicity, educational attainment, what partner was doing last week, date last worked for two consecutive weeks, health, substance abuse, number of biological children, interaction with child (activities, discipline), R’s rapport with partner, abuse, incarceration history. Section F: Demographics Size of household, household roster (name, gender, age, relationship to R, employment status), number of children living outside household. Section S: Word Association Section H: Mother’s Family Background and Support Birthplace of father/mother, citizenship, educational attainment of father/mother, family structure, financial assistance from friends/family, extent of R’s support group, how often does child see R’s parents, how often does child see FATHER’S parents. Section I: Environment and Programs Involvement and/or participation in various groups and organizations, registered/eligible voter, view of the importance of rights of US citizenry, current housing situation (house/apartment/condo, rent/mortgage amount, public housing), lack of utilities due to insufficient funds, assistance received from government agencies (i.e. Head Start, WIC, AFDC/TANF, Food Stamps, Unemployment Insurance, Workers’ Compensation, SSI, EITC, etc.), other problems due to lack of funds (i.e. need to seek free food, eviction, borrow money from friends, forgone medical treatment, etc.), legal charges pending, incarceration history. Section J: Health and Health Behavior Satisfaction with life, opinion of health status, presence of health problems (regular medication), hospital visits, height/weight, FATHER’S height/weight, medical coverage (state, federal or private) medical coverage for child, depression (presence, duration, intensity, side affects), anxiety (presence, duration, intensity, side affects), alcohol use/abuse, drug use/abuse, counseling and hospitalizations, views of own behavior, parents’ history with alcohol, drugs, and mental illness. Abt Associates Inc. Chapter 3: The Fragile Families and Child Wellbeing Study 27 Exhibit 3.5 Fragile Families: Content of Three-Year Follow-Up Mothers’ Instrument Section R: Religion Affiliation, attendance, views of religious texts, transforming/religious experiences. Section K: Education and Employment Current educational activity, educational attainment, job search, reservation wage, date last worked for two consecutive weeks, earnings, work hours, information on second job, information on “off-the-books” work, number of regular jobs in last three years, own business, illegal activities for money. Section L: Income Total household income, money given or loaned to friends, bank account (primary holder), credit card (primary holder), who handles financial responsibilities, car/truck/van (amount owed, resale value), what does R enjoy doing with child the most. a FATHER refers to biological father of the focal child. 28 Chapter 3: The Fragile Families and Child Wellbeing Study Abt Associates Inc. Exhibit 3.6 Fragile Families: Content of Three-Year Follow-Up Fathers’ Instrument Section A: Family Characteristics Child’s current living situation, Father’s visitation with non-residential children, relationship with MOTHER,a date of MOTHER’S death, date of marriage to MOTHER, start date of cohabitation with MOTHER, date of divorce from MOTHER, date romantic relationship with MOTHER ended, reason for relationship termination, current relationship with MOTHER, subsequent children since birth of focal child, number of children with MOTHER, number of children with different mothers, number of romantic relationships before MOTHER, pregnancies in these relationships and outcomes. Section B: Child Wellbeing and Fathering R’s view of child’s health and development, separations between R and child (dates, child care during separation, reasons for separation), interaction with child (activities, discipline), views on parenthood, childcare arrangements, money/voucher/scholarship for childcare. Section C: Mother-Child Relationship Opinion of MOTHER’s relationship with child, interaction between MOTHER and child (frequency, activities, discipline), legal arrangements with MOTHER (paternity, child support, visitation), informal child support agreements, MOTHER’S support of children other than focal child, MOTHER’S current marital status, MOTHER’S current residence, is MOTHER currently in jail (offense), MOTHER’S health, MOTHER’S substance abuse. Section D: Father’s Relationship with MOTHER (for fathers who are or were in a relationship) Opinion of MOTHER as a mother, R’s trust with MOTHER, rapport with MOTHER, frequency of certain behaviors, abuse. Section E: Current Partner Gender, marital status, date of marriage, date cohabitation began, length of romantic involvement, date of birth, race/ethnicity, educational attainment, what partner was doing last week, date last worked two consecutive weeks, health, substance abuse, number of biological children, interaction with child (activities, discipline). Section F: Demographics Size of household, household roster (name, gender, age, relationship to R, employment status), number of children living outside household. Section S: Word Association Section H: Father’s Family Background and Support Date of birth, race/ethnicity, birthplace of father/mother, citizenship, educational attainment of father/mother, family structure, financial assistance from friends/family, extent of R’s support group, how often does child see R’s parents, how often does child see MOTHER’S parents. Section I: Environment and Programs Involvement and/or participation in various groups and organizations, registered/eligible voter, view of the importance of rights of US citizenry, current housing situation (house/apartment/condo, rent/mortgage amount, public housing), lack of utilities due to insufficient funds, assistance received from government agencies (i.e. Head Start, WIC, AFDC/TANF, Food Stamps, Unemployment Insurance, Workers’ Compensation, SSI, EITC, etc.), other problems due to lack of funds (i.e. need to seek free food, eviction, borrow money from friends, forgone medical treatment, etc.), legal charges pending, incarceration history. Section J: Health and Health Behavior Satisfaction with life, opinion of health status, presence of health problems (regular medication(s)), hospital visits, height/weight, MOTHER’S height/weight, medical coverage (state, federal or private), medical coverage for child, depression (presence, duration, intensity, side affects), anxiety (presence, duration, intensity, side affects), alcohol use/abuse, drug use/abuse, counseling and hospitalizations, views of own behavior, parents’ history with alcohol, drugs, and mental illness. Section R: Religion Affiliation, attendance, views of religious texts, transforming/religious experiences. Abt Associates Inc. Chapter 3: The Fragile Families and Child Wellbeing Study 29 Exhibit 3.6 Fragile Families: Content of Three-Year Follow-Up Fathers’ Instrument Section K: Education and Employment Current educational activity, educational attainment, job search, reservation wage, date last worked for 2 consecutive weeks, earnings, work hours, information on second job, information on “off-the-books” work, number of regular jobs in last three years, own business, illegal activities for money. Section L: Income Total household income, money given or loaned to friends, bank account (primary holder), credit card (primary holder), who handles financial responsibilities, car/truck/van (amount owed, resale value), what does R enjoy doing with child the most. a MOTHER refers to biological mother of the focal child. Exhibit 3.7 Fragile Families: Attitudinal Items on Gender Roles • • • • • • The important decisions in the family should be made by the man of the house. It is much better for everyone if the man earns the main living and the woman takes care of the home and family. In a dating relationship, a man [woman] is largely out to take advantage of a woman [man]. Men [Women] cannot be trusted to be faithful. It is more important for a man to spend time with his family than to work as many hours as he can. Fathers play a more important role in raising boys than in raising girls. Note: Bracketed words indicate adaptation of the items for fathers versus mothers. Exhibit 3.8 Fragile Families: General Attitudes Towards Marriage, Keys to a Successful Marriage General attitudes toward marriage: • The main advantage of marriage it that it gives financial security. • All in all, there are more advantages to being single than to being married. • A mother living along can bring up her child as well as a married couple. • It is better for a couple to get married than just to live together. • It is better for children if their parents are married. • Living together is just he same as being married. How important do you think the following qualities are for a successful marriage? • Having the same friends • The husband having a steady job • Being of the same race/ethnic group • Having good sex • Having the same religious beliefs • The husband and wife being emotionally mature 30 Chapter 3: The Fragile Families and Child Wellbeing Study Abt Associates Inc. Exhibit 3.9 Fragile Families : Attitudes Towards Fatherhood Fathers do many things for their children. Please tell me how important each of the following activities is to you: • Provide regular financial support • Teach child about life • Provide direct care, such as feeding, dressing and child care • Show love and affection to the child • Provide protection for the child • Serve as an authority figure and discipline the child Which of these is the most important to you for a father to do for his child(ren)? • Provide regular financial support • Teach child about life • Provide direct care, such as feeding, dressing and child care • Show love and affection to the child • Provide protection for the child • Serve as an authority figure and discipline the child Which of these is the least important to you for a father to do for his child(ren)? • Provide regular financial support • Teach child about life • Provide direct care, such as feeding, dressing and child care • Show love and affection to the child • Provide protection for the child • Serve as an authority figure and discipline the child Now I’d like to ask you some questions about the rights and obligations of fathers who live apart from their children. If a father provides financial support to the mother, do you think he should have a right to: • See his child on a regular basis • Make decisions about how the child is raised If a father can afford to provide financial support but does not, do you think he should have the right to: • See his child on a regular basis • Make decisions about how the child is raised If a father cannot afford to provide financial support, do you think he should have the right to: • See his child on a regular basis • Make decisions about how the child is raised If a mother has a new partner, do you think the baby’s father should be required to provide financial support to his baby? If the father has a new baby with another partner, do you think he should be required to provide support to his children from a previous relationship? Abt Associates Inc. Chapter 3: The Fragile Families and Child Wellbeing Study 31 Chapter Four: The National Longitudinal Survey Of Youth The National Longitudinal Survey of Youth (NLSY) is part of the NLS (National Longitudinal Surveys), a set of surveys that collects information on the labor market experiences of various subgroups of the population. The NLSY has several components, corresponding to different cohorts. The main components are the NLSY79, which follows a nationally representative sample of (originally) young people who were aged 14 to 21 at the end of 1978; and the NLSY97, which follows a nationally representative of young people who were aged 12 to 16 at the end of 1996. The primary purpose of both the NLSY79 and the NLSY 97 is the collection of data on respondents’ labor force participation and investments in education and training. Other core topics are marital history, fertility, health, and income and assets. Periodic topical modules have covered a wide variety of other topics. The NLSY is sponsored by the Bureau of Labor Statistics, in the US Department of Labor. Supportive funding is provided by the U.S. Department of Health and Human Services, NICHHS, National Institute on Aging, National Institute on Alcohol Abuse and Alcoholism, and the National Institute on Drug Abuse. Additional funding has been provided by the US Department of Defense and the Armed Services, and by the National Institute of Education. Survey Design The NLSY79 sample consists of three parts: a nationally representative sample of noninstitutionalized civilians aged 14 to 21 as of December 31, 1978 (n=6,111), a supplemental sample of civilian Hispanic, black, and economically disadvantaged youth (n=5,295); and a sample of youth enlisted in the military (n=1,280). The sampling design led to the inclusion of a large number of sibling pairs: 5,863 civilian respondents had at least one sibling included in the survey. The poor, non-black non-Hispanic subsample was dropped after the 1990 interview. The survey was administered annually through 1994, and biennially since then. An offshoot of the NLSY79, the NLSY Children and Young Adults Survey, began in 1986. It initially comprised all children born to NLSY79 female respondents. Starting in 1988, limited information has been collected biennially from children aged 10 or older. Since 1994, in-depth interviews of youth 15 and older have been conducted biennially, following the NLSY79 questionnaire. The sample only covers children of women currently aged 36 to 43, and is thus not representative of all children. The sample frame also excludes offspring of women who immigrated to the United States after 1979. The sample size as of 1998 was 10,918 individuals, including 2,143 who were between the ages of 15 and 20. (Children aged 21 and over were not interviewed in 1998.) These data are of greatest interest for analyzing child development relative to maternal and other background characteristics. The NLSY97 sample similarly consists of two parts: a nationally representative sample of youths born between 1980 and 1984 and a supplemental sample of black and Hispanic youths. All appropriately aged individuals in selected households were included. In the first round, a parent of each youth was also interviewed. The survey is annual. Abt Associates Inc. Chapter 4: The National Longitudinal Survey Of Youth 33 Sample sizes: The initial sample sizes for NLSY79 and NLSY97 were 12,696 and 9,022, respectively. Methods of data collection: NLSY79 interviews have generally been conducted in person. In 1987, a shorter interview was administered by telephone due to budget constraints. CAPI instruments were introduced in 1993. NLSY97 uses CAPI for most of the survey instrument and A-CASI (audio computer-assisted selfinterview) to gather potentially sensitive information from the youth (i.e. drug use, sexual behavior) and, in Round 1, from the youth’s parents. Response rates: As of 1998, NLSY79 had 8,399 respondents (out of an original sample size of 12,686). Response rates in 1998 and 2000 were around 85 percent. The NLSY97 retention rates for 1998/1999, 1999/2000, and 2000/2001 were 93.3 percent, 91.4 percent, and 89.9 percent, respectively. Survey Content NSLY79 administers a core questionnaire, a series of topical modules, and other special data collections. These latter included the 1980 high school survey, the 1980-83 high school transcript data collections, collection of standardized test scores, and the 1980 administration of the Armed Services Vocational Aptitude Battery (ASVAB). The core topics for NLSY79 are marital and fertility histories, work and non-work experiences, training investments, schooling, family income and assets, health conditions and injuries, geographic residence and environmental characteristics, and household composition. Topical modules have included family background, knowledge of the work world, retrospective labor market experience, job search methods, migration, school discipline, health, childcare, self-esteem, locus of control, depression, time use, delinquency, knowledge about AIDS, childhood residences, neighborhood problems, drug and alcohol use, and prenatal and postnatal health behaviors. Some supplements of special interest for the current topic have collected data on survey members’ attitudes toward women and work, educational and occupational aspirations, future expectations about marriage, and the attitude of the most influential person in their life toward childbearing and other decisions. The NLSY97 likewise comprises a core questionnaire, topical modules, and a number of special data collections. One of these was an interview with a parent conducted during Round 1, covering the parent’s marital and employment history and other background information. Another was a survey of school administrators, on characteristics of schools in the youth’s community: total enrollment, grades served, average daily attendance, racial/ethnic composition, programs available, number of reported incidents of gang activity and vandalism, weapons possession and alcohol or drug abuse, and others. The school survey has occurred twice, in the winters of 1996 and 2001. The Armed Services Vocational Aptitude Battery (ASVAB) was also administered to sample members in a computeradapted form. 34 Chapter 3: The Fragile Families and Child Wellbeing Study Abt Associates Inc. A household roster collected information on all members of the respondent’s household, and the nonresident roster gathered information on those members of the respondent’s immediate family who lived elsewhere. For each person, these rosters collected demographic information, marital status, educational attainment and employment status. These were completed in Round 1 of data collection and updated in subsequent rounds. The core questionnaire of NLSY97 comprises the following sections: household information, schooling, employment, training, time use, health, marriage, fertility, program participation, income, and assets. A self-administered section covers the youth’s relationship with parents, sexual activities, use of cigarettes, alcohol, and drugs, and illegal activities. Suitability for Research Outcome measures: The surveys collect ongoing information on union status, formation, and dissolution. Information on cohabitation has been included in the NLSY since 1990, including when an unmarried couple started living together, when a married couple started living together, and whether a respondent and spouse lived together continuously until marriage. Cohabitational and marital union quality is measured in the core questionnaire by items on satisfaction, frequency of various interactions (calm discussion, laughing together, describing one’s day), and frequency of arguments on various subjects (e.g. chores, children, money). Measures of demographic influences: NLSY79 is an appropriate data source for investigating key demographic influences on union behaviors: teen and non-marital childbearing, transitions to parenthood, and effects of cohabitation on marriage. This research is possible due to the collection of full fertility and marital histories (including information about cohabitation). Some supplements have collected information on positive and negative interactions in relationships and some relevant attitudinal data. NLSY97 offers the opportunity to examine young marriages, as well as the short-term influence of teen childbearing on union formation and quality with a timely and thus policy relevant data source. Measures of economic influences: NLSY is an appropriate data source for investigating key economic issues relevant to union behaviors. It has extensive and highly detailed measures of both male and female labor market experiences. NLSY collects data on the characteristics of up to five jobs per year: start and stop dates, hours worked, pay rates, occupation, industry, class of employer (i.e. government, private for-profit company, not-for-profit company, self-employed, family business), reason for leaving, and work gaps occurring between the start and stop dates for a given employer. NLSY79 also includes unusual information on knowledge of the work world, attitudes towards work, self-esteem, and vocational ability (the ASVAB). Geocoded data are available by special request to allow linkages to measures of the local economic environment, providing information about economic opportunities. These same geocoded data could be utilized in studies of the marriage market, allowing aggregate level measures of the quality and quantity of potential partners to be linked to the individual level data. Measures of socio-cultural influences: The NLSY has been used in many past sociological studies of union behaviors. It is well-suited to examining the influence of gender role attitudes on union Abt Associates Inc. Chapter 4: The National Longitudinal Survey Of Youth 35 behaviors. Eight items about women’s roles were included in three rounds of NLSY79 (1979, 1982, 1987; Exhibit 4.3). (These have not been collected in NLSY97.) The longitudinal nature of the data collection has two primary advantages for this area of investigation. First, the attitudinal measures were collected in the first year of the survey, permitting proper temporal ordering for drawing causal conclusions about the influence of preexisting attitudes on later behaviors. Second, since the gender role attitude questions were collected at three points in time, there is an opportunity to examine individual changes in these attitudes, and the influence of union experiences on these attitudes. An additional strength is that these questions were collected directly from men and women, allowing comparison of differences by gender in the association between the attitudes and union behaviors. Unfortunately , the weakness of these measures is that they were not collected after 1987. Given the ongoing shifts in family attitudes in the U.S., the NLSY may offer an outdated perspective on gender role attitudes. Another shortcoming is that the NLSY does not collect data on attitudes regarding the meaning of marriage and cohabitation, which would be useful for better understanding of socio-cultural influences on union behaviors. Measures of psychological influences: NLSY79 has some data on behaviors and relationship satisfaction, but not from both members of a couple. Topical modules address several personal traits: self-esteem, sociability, locus of control, and depression. Special strengths: The NLSY79 is a powerful data source for tracing the coming of age of young people in the 1980s and the sequelae. Collecting data directly from both men and women, it allows linking of their subsequent marital history over the next two decades and more to their backgrounds and labor market experiences. Unusually ric h data are available on the sample’s education, cognitive ability, and childhood living arrangements. The breadth of potential covariates relevant to union behaviors is impressive, spanning the range from economic and labor market information, attitudes and expectations, family and household issues, to cognitive and psychological measures. The inclusion of sibling pairs in the sample supports estimation of fixed and random effects models. The NLSY97 promises to be another equally valuable data source, as similarly rich data are being collected. Limitations: The NLSY79 sample is too old for studying marital formation in general in the 1990s; by 1994, respondents were already aged 30 to 37. Conversely, the NLSY97 sample is probably still too young (alt hough getting older all the time). In 2000, they were aged 16 to 20. Using the Data The website for the NLSY is http://www.bls.gov/nls/home.htm. Documentation: Technical documentation is available on line at http://www.bls.gov/nls/nlsdoc.htm. Available reports include the NLSY97 User’s Guide (2001), the NLSY79 User’s Guide (2000), and the NLSY79 Child/Young Adult User’s Guide (1998). Questionnaires and codebooks may also be downloaded. Data availability: Data are available both on CDs for a nominal price ($20.00 for Rounds 1 through 4 of NLSY97, $20.00 for 1979-2000 NLSY79) and as free downloads. 36 Chapter 3: The Fragile Families and Child Wellbeing Study Abt Associates Inc. Reports and articles: The NLS website includes a bibliography of over 3000 publications. Some examples of reports on this topic that use the NLSY are: McLaughlin, D. K., and Lichter, D. T. (1997). Poverty and the marital behavior of young women. Journal of Marriage and the Family, 59, 582-594. Oppenheimer, V. K., Kalmijn, M., and Lim, N. (1997). Men's career development and marriage timing during a period of rising inequality. Demography, 34(3), 311-330. Sweeney, M. M. (2002). Two decades of family change: The shifting economic foundations of marriage. American Sociological Review, 67, 132-147. Data Elements The core questionnaires for NLSY79 and NLSY97 are described in Exhibits 4.1 and 4.2, respectively. Gender role items from NLSY79 are listed in Exhibit 4.3. Abt Associates Inc. Chapter 4: The National Longitudinal Survey Of Youth 37 Exhibit 4.1 NLSY79: Content of Round 19 Questionnaire (2000) Household Interview Address, household roster (name, gender, relationship to R, age, employment status, educational attainment), identification of part-time household members, migratory activity, religious affiliation and attendance. Marital History Marital status (current, at last interview, changes since last interview), date of (re)marriage, R’s most recent spouse (date of birth, cohabitation before marriage, religious affiliation and attendance, marital history), R’s most recent partner (date of birth, date cohabitation began, religious affiliation and attendance, marital history). Schooling Educational activity since last interview, current educational activity, educational attainment, date of GED/diploma/degree, post-secondary education (institution, location, dates enrolled, earned credit hours, student loans). Military Military service since date of last interview. Employment Employer(s) at last interview (name, date left, reason for leaving), employer(s) prior to last interview (name, date resumed working, date left), work for pay since last interview (name(s) of employer(s), job activities/occupation, date employment began/ended. Employer Supplement Information on employer(s) since date of last interview. Start date, number of months worked, reason for leaving; presence, amount, and duration of severance pay; dates of employment gaps, reasons for gaps; job search(es) (length, reason for not looking, methods, job offers), maternity leave (date began/ended), hours worked per week, activities/duties, type of business/industry, occupation, number of employees, earnings, changes in earnings (date, reason), benefits received/available. Gaps in Employment Job search (weeks searching, reason for not looking, methods used, job offers (number, offered wage, reason for declining), reason for leaving last job, date last employed. Training Style, how was training paid for (i.e. employer, family, loans, etc.), who ran the training, prerequisites (tenure with given employer), date training started, date training was completed/date left training, was training completed, hours per week, reason for training, skills acquired, informal, on-the-job training. Spouse Labor Force Activity Change in usual activities and duties since last interview, occupation, usual duties and activities, earnings, hours per week, number of weeks worked at all jobs. Fertility Gender and date of birth for biological and non-biological children; offspring since last interview: number, name, gender, date of birth, child’s residence, residence of other parent and when other parent left household, frequency of other parent’s visits, child’s health; fertility expectations, methods of birth control used in last month, relationship with children; pregnancies since 1998: number, outcome, date of outcome, behaviors during pregnancy, child’s health history for 1st year, well baby care, breast feeding history; grade level of school-age children. 38 Chapter 3: The Fragile Families and Child Wellbeing Study Abt Associates Inc. Exhibit 4.1 NLSY79: Content of Round 19 Questionnaire (2000) Child Care Questions are for child care used in child’s first year of life, first three years, second year of life and third year of life (for up to three arrangements. Number of child care arrangements used for 1+ month(s), location, date child care arrangement began, how long did arrangement last. Questions about current marriage/relationship Satisfaction, frequency of various interactions, subject of arguments. Other topics Opinion of neighborhood as a place to raise children, problems with neighborhood. Health Health limitations to work: date limitation began, health requirements of job(s), frequency of participation in physical exercise, sports, or heavy housework; on-job illnesses or injuries: employer, date of occurrence, activity at time of illness/injury, body part(s) hurt or affected, nature of illness/injury, work limitations due to illness/injury, duration of illness/injury, consequences of illness/injury (i.e. laid off, fired, quit, etc.; health care and hospitalization plans: policy source, plan type, characteristics of plan, months not covered in past year, date last covered by insurance; spouse/partner’s insurance coverage, self-assessment of R’s health, most recent visit to a health care professional, health of biological parents, presence, date diagnosed, and limitations of various health problems. Income and Assets (for 1999) Income sources and amounts, income sources and amounts for R’s spouse/partner, child support amount and terms of agreement, child support obligations, debt sources and amounts, assets and their estimated value. Abt Associates Inc. Chapter 4: The National Longitudinal Survey Of Youth 39 Exhibit 4.2 NLSY97: Content of Round 3 Questionnaire (2000) Household Information Changes in residence since last interview: number, addresses, date of move; current housing type; living situation: does R live with parent(s), which parent, custodial arrangement, name of other custodian, primary residence, dates lived with parent, reason for not living with other parent; non-resident parent: name, relationship to R, does R live with non-residential parent at all, dates lived with non-residential parent; parents’ marital status, parental figure: name, relationship to R, gender does R live with parental figure at all; household roster (name, gender, age, race/ethnicity, religious preference, marital status, educational attainment, relationship to R, age). Schooling Current educational activity, educational history: educational attainment at time of last interview, school name/location/type, dates of enrollment, reason for leaving, history of school suspensions, school marks in 8th grade/high school, standardized testing: tests taken, grade when taken, scores, AP testing: subjects taken, high school: program of study, courses taken, date received GED/diploma, name and location school, college: name of college(s), credits, GPA, major(s), tuition and fees, financial aid, School-to-work program participation. Employment Employer(s) at last interview: name, start date, date left, reason for leaving, new employers since last interview: name, start date, date left, reason for leaving; characteristics of current/most recent employment: employer characteristics, job activities/occupation, earnings, hours per week, benefits, gaps in employment: dates, reason for gap, job search, self-employment/freelance: type of work, start date, hours per week, earnings, activities/duties. Training Excluding regular schooling. Training program(s): date of attendance, type of program, government aid/provision, program location, means of payment, reason for enrollment, means of measuring progress, certificate/license received, reason for noncompletion, skills acquired. Time Use Weekdays: time R wakes up, leaves house, arrives home, goes to bed, typical weekday activities. Health Opinion of own health, approximate height, approximate weight. Self-Administered View of home life, relationship with family, relationship with mother/father, contact/relationship with nonresidential, biological parent(s), relationship with adopted mother/father, parental discipline/control, relationship between parents/parent figures/ biological parents, puberty, dating history: age started dating, number of dates since last interview, incidence of sexual intercourse (age of first intercourse, “safe sex” practices, number of partners, incidence of pregnancy and outcomes, tobacco, alcohol, marijuana, and other drug use, history of running away, history with firearms, gang experience, involvement in illegal activities, incarceration history). Marriage Current marital status, date cohabitation began, spouse(s)/partner(s) since last interview: age at time cohabitation began, educational attainment, religious affiliation, race/ethnicity, date relationship ended. Fertility Biological children at time of last interview: name, gender, date of birth, living situation; biological children since last interview: name, gender, date of birth, living situation; other parent of children: name, race/ethnicity, age, educational attainment, religious affiliation, relationship with R; adopted children: name, gender, date of birth, living situation. 40 Chapter 3: The Fragile Families and Child Wellbeing Study Abt Associates Inc. Exhibit 4.2 NLSY97: Content of Round 3 Questionnaire (2000) Program Participation For respondent and respondent’s spouse/partner. Unemployment compensation, workers’ compensation, food stamps, WIC, public housing, AFDC or ADC, SSI and other welfare programs. Income (for 1998) Income sources and amounts, income sources and amounts for R’s spouse/partner, earnings for R’s biological mother, father, non-biological mother/female guardian, non-biological father/male guardian, other persons over age 14 in household, spouse/partner’s earnings during 1997. Assets Assets and their estimated value, debt sources and amounts. PIAT Math Assessment Exhibit 4.3 NLSY79: Attitudes on Gender Roles General attitudes toward marriage: • Woman’s place is in the home, not the office or shop. • A wife with a family has no time for outside employment. • A working wife feels more useful than one who doesn’t hold a job. • Employment of wives leads to more juvenile delinquency. • Employment of both parents is necessary to keep up with the high cost of living. • It is much better if the man is the achiever outside the home and the woman takes care of the home and family. • Men should share the work around the house with women. • Women are much happier if they stay home and take care of children. Abt Associates Inc. Chapter 4: The National Longitudinal Survey Of Youth 41 Chapter Five: The National Survey Of America’s Families The National Survey of America’s Families (NSAF) is part of “Assessing the New Federalism,” a research project being carried out by the Urban Institute in partnership with Child Trends. Funding is provided by a consortium comprising the Annie E. Casey Foundation, the W.K. Kellogg Foundation, the Robert Wood Johnson Foundation, the Henry J. Kaiser Family Foundation, the Ford Foundation, the David and Lucile Packard Foundation, the John D. and Catherine T. MacArthur Foundation, the Charles Stewart Mott Foundation, the McKnight Foundation, the Commonwealth Fund, the Stuart Foundation, the Weingart Foundation, the Fund for New Jersey, the Lynde and Harry Bradley Foundation, the Joyce Foundation, and the Rockefeller Foundation. The purpose of the study is to analyze how the well-being of children and families has been affected by PRWORA through the devolution of respons ibility for social programs to the States. To support this analysis, data are being collected on children in low-income households (under 200 percent of poverty), and also, for purposes of comparison, on children in higher income households, on lowincome adults under age 65, and on other adults aged 65 and younger. Survey Design NSAF is a repeated cross-sectional survey, with three rounds of data collection, in 1997, 1999, and 2002. Each round corresponds to a new sample of respondents. The samples are nationally representative, and are also representative for the District of Columbia and 12 states: Alabama, California, Colorado, Florida, Massachusetts, Minnesota, Mississippi, New York, New Jersey, Texas, Washington, and Wisconsin. The survey begins with a screener to determine whether households should be administered an extended interview, tailored to whether children under 18 are present. The “most knowledgeable adult” (MKA) is interviewed about one or two children—a randomly selected child from those under age 6, if any, and another randomly selected child from those aged 6 to 17, if any. The MKA also provides information about his or her own spouse. Sample sizes: In both 1997 and 1999, about 30,000 interviews were conducted with primary caregivers of children. An additional 20,000 interviews in 1997 and 17,000 in 1999 were conducted with childless adults.12 The sample sizes in the 13 selected states are sufficient for state-level analysis. For example, in 1997 the number of sampled children in the various states ranged from 1,984 to 2,566, and the number of sampled low-income children from 950 to 1,366. Methods of data collection: All interviews are conducted via CATI, primarily through a random digit dialing (RDD) survey. A supplementary survey is conducted of households without telephones to ensure full coverage of low-income families. These households are given cell phones to allow them to call in to the interviewing center and to be surveyed via CATI as well. 12 The exact counts for 1997 and 1999 respectively were 28,331 and 29,917 interviews with caregivers, and 20,168 and 16,788 interviews with childless adults. Chapter 5: The National Survey Of America’s Families 43 Abt Associates Inc. Response rates: In 1997, the response rates were 82 percent for the MKA of sample children, 69 percent for other adults in households with children, 77 percent for adults in adult-only households, and over 80 percent for the non-telephone sample. Survey Content The NSAF survey has changed only slightly over the three rounds of data collection. The topics covered are economic security, including employment and participation in social welfare programs, health and health care, child’s education and cognitive development, child’s social development and positive development, child’s behavior problems, family environment, and community environment. The closing section collects some respondent views on welfare and on gender roles. Suitability for Research Outcome measures: Marital and cohabitational status at the time of interview is recorded, but not union history. There are no items on union quality. Measures of demographic influences: NSAF’s sample characteristics, with its emphasis on low- income families with children, make it appropriate for some analyses of union behaviors related to teen and non-marital childbearing. But NSAF is far from ideal for studying these topics. The lack of marital, cohabitational, and fertility histories and the purely cross-sectional nature of the data are key drawbacks. Additionally, although NSAF contains several unusual items on respondents’ attitudes towards out-of wedlock childbearing, the attitudinal information is not prospective relative to union and childbearing status, so that causation cannot be attributed. The structure and content of the survey would not support analyses of transitions to parenthood or effects of cohabitation on marriage. Measures of economic influences: A strength of the NSAF is that good information is collected on earnings and employment of respondents and their spouses and partners, all relevant for exploring economic influences on union behaviors. However, the cross-sectional nature of the data is again problematic, and these economic measures are not prospective relative to union status. Attribution of causation would therefore be highly questionable. Measures of socio-cultural influences: NSAF includes several attitudinal measures on the meaning of marriage and on gender roles which would be interesting to examine across different subgroups of the population, especially relative to disadvantage. Such descriptive studies might offer insights relevant to union behaviors. There is the potential to test for aggregate change in these attitudes across the three repeated cross-sections, although it seems unlikely attitudes could have shifted too sharply in the relatively small window of time covered. The weakness is that it would be difficult to use NSAF to directly examine the influence of these attitudes on union behaviors, as the attitudes are measured simultaneously with current union status. Measures of psychological influences: NSAF does not collect information on behaviors, processes, or relationship satisfaction. Special strengths: The NSAF offers large sample sizes, especially of low-income households with children, and permits analysis in many individual states and the District of Columbia. It collects 44 Chapter 5: The National Survey Of America’s Families Abt Associates Inc. substantial information on employment and earnings and other income sources. It includes measures of interaction between children and their non-custodial parents, and also a few attitudinal measures. The repeated cross-sectional design encourages analyses of trends over time. NSAF derives additional strength for understanding the influence of welfare reform on the well-being of low income families by links to the Urban Institute’s other related data collection efforts. NSAF is one component of a larger research initiative, “Assessing the New Federalism.” Other aspects of this project include detailed reports of policy initiatives and programs related to PRWORA in the District of Columbia and each of the 12 core states. Additionally, the Urban Institute has made available on the internet a 50-state database of policy measures potentially relevant to family well-being. These other research components offer the opportunity to link individual level information from NSAF to detailed information about the state policy context, relating union status to the incentives and disincentives provided by means-tested programs. Limitations: The NSAF is focused primarily on the well-being of children and families. Hence it does not sample low-income childless adults, a population of interest for studying union formation. It interviews each household only once, and does not collect any information on marital or cohabitational history, or on union quality. Because of these design constraints, NSAF offers only limited opportunities to investigate the causes of union formation, quality or dissolution. Instead, NSAF probably is most appropriate for description of differences in the experiences of individuals related to their current union status, and variations in these union-specific experiences related to PRWORA. Using the Data The NSAF data are described on the Urban Institute’s website, at http://newfederalism.urban.org. Documentation: A series of reports on the sample design, the questionnaire, survey weights, survey methods, and other topics is available online at: http://www.urban.org/content/Research/NewFederalism/NSAF/Methodology/ 1999MethodologySeries/1999.htm Data availability: Public use data files are available for the first two rounds of data collection. Data files for the third wave of data collection (currently in the field) are expected to be available in 2004. Date files are available free of charge for downloading from the New Federalism website, but users are required to register on-line. Reports and articles: A list of Urban Institute and Child Trends reports using NSAF may be found at: http://www.urban.org/content/Research/NewFederalism/Publications/PublicationsbyTopic /ByTopic.htm. These include, for example: Abt Associates Inc. Chapter 5: The National Survey Of America’s Families 45 Acs, G. & Nelson, S. (2002). Assessing the relationship between welfare policies and changes in living arrangements of low-income families and children in the late 1990s. Discussion Paper 02-05. Lerman, R. I. (2002). How do marriage, cohabitation, and single parenthood affect the material hardships of families with children? Urban Institute, (http://www.urban.org/url.cfm?ID=410539). Data Elements The most recent version of the NSAF questionnaire is described in Exhibit 5.1. Attitudinal items regarding welfare and gender roles are shown in Exhibit 5.2. 46 Chapter 5: The National Survey Of America’s Families Abt Associates Inc. Exhibit 5.1 NSAF: Content of the 2002 Questionnaire Section B: Health Status and Satisfaction Satisfaction with current medical care, child’s health status, knowledge of federal/state-provided health care. Section C: Child Educationa (for two children) Child’s current grade level/last grade completed, disciplinary problems in school, child’s employment. Section D: Household Roster Age, gender, usual residence, marital status, relationship to R, relationship to child, presence of child’s mother/father, relationship to R’s spouse/partner, identification of most knowledgeable adult (MKA). Section E: Health Coverage Policy holder, persons covered under plan, presence of additional plans (policy holder), presence/identification of persons covered by Medicare, presence/identification of persons covered by other government-sponsored plans (i.e. CHAMPUS, TRICARE, CHAMP-VA, other military health care, or the Indian Health Service), presence/identification of persons covered by state-sponsored plans, characteristics of health insurance plans, identification of persons without health care coverage, health insurance coverage history. Section F: Health Care Use and Access Health status of R/R’s spouse/partner, presence of supervision or special care for R/R’s spouse/partner, dental visits for R/R’s spouse/partner/child, mental health services for R/R’s spouse/partner/child, well-child check-ups, characteristics of regular check-ups, difficulties with health care in last twelve months, amount spent on health care in last twelve months. Section G: Child Care Means of child care (i.e. Head Start, school-provided program, babysitting, etc.), R’s activities while child was in care, characteristics of child care provider, amount spent on child care, presence of child care a id. Section H: Non-Residential Parent Interaction between child and parent(s), child support. Section I: Employment and Earnings (for R and R’s spouse/partner) Employment history, current employment status, characteristics of current employment, earn ings, employment in 2001. Section J: Family Income Presence, amounts, and recipients of income from AFDC/TANF, state or county programs, child support, foster care payments, financial assistance from friends or family, unemployment insurance, workers’ c ompensation, VA benefits, SSI, SSDI, private disability insurance, Social Security, pensions, interest on assets, rental property, or other. Section K: Welfare Program Participation Current status and history with welfare programs. Section L: Education and Training (for R and R’s spouse/partner) Educational attainment, job training during 1998. Section M: Housing and Economic Hardship Current living situation (rent/mortgage amount, public housing, housing characteristics, etc.), lack of food/receipt of food assistance, lack of utilities. Section N: Issues, Problems, Social Services Frequency of certain feelings, frequency of certain behaviors in children, children’s participation in various programs, R’s participation in certain programs, attendance of religious services. Section O: Race, Ethnicity, and Nativity (for members of R’s household) Race/ethnicity, citizenship, year arrived in US. Section P: Closing Section Opinions of welfare and working, opinions on raising children, R’s address. a There are two versions of section C: a main version and a summer version if survey was conducted during the summer months. Abt Associates Inc. Chapter 5: The National Survey Of America’s Families 47 Exhibit 5.2 NSAF: Attitudinal Items Strongly agree/agree/disagree/strongly disagree: P1. Here are some opinions that people have expressed about welfare and about working a a. Welfare makes people work less than they would if there wasn’t a welfare system. b. Welfare helps people get on their feet when facing difficult situations such as unemployment, a divorce, or a death in the family. c. Welfare encourages young women to have babies before marriage. P2. The following are some opinions that others have expressed about raising children: a. A single mother can bring up a child as well as a married mother. b. A working mother can establish just as warm and secure a relationship with her children as a mother who does not work. c. People who want children ought to get married. d. When children are young, mothers should not work outside the home. a A fourth item, about t he importance of work, was deleted after 1997 because virtually all respondents agreed or strongly agreed. 48 Chapter 5: The National Survey Of America’s Families Abt Associates Inc. Chapter Six: The National Survey Of Family Growth The National Survey of Family Growth (NSFG) is an ongoing project, currently entering its sixth wave or cycle of data collection. Previous cycles of data were collected in 1973, 1976, 1982, 1988, and 1995. The 2002 NSFG, referred to as Cycle 6, is currently in the field and it is anticipated that public use data will be available sometime in 2003. The main purpose of the NSFG is to provide reliable national data on marriage, divorce, contraception, fertility, infertility, and the health of women and infants in the United States. Research based on the NSFG covers topics such as sexual activity, contraceptive use, the effectiveness of contraceptives, cohabitation, marriage and divorce, childbearing, infertility, and use of family planning and other health services. The NSFG is sponsored by the National Center for Health Statistics (NCHS), in the Centers for Disease Control and Prevention (CDC). Major funding for 1982, 1988, and 1995 came from three parts of the Department of Health and Human Services (DHHS): the NCHS, the Office of Population Affairs, and the National Institute for Child Health and Human Development. Survey Design The NSFG is a repeated cross-sectional survey. Its sample frame has changed significantly over its history. For the first five cycles (1973-1995) the NSFG interviewed women aged 15-44 years in the civilian, non-institutionalized population. Cycles 1 and 2 (1973 and 1976) were restricted to women who were currently or formerly married. Subsequently, the NSFG was redesigned to include women who had never been married. For Cycle 5 (1995), a sample of women aged 15 to 44 was selected from households that responded to the 1993 National Health Interview Survey (NHIS). The NHIS is a continuous survey of the US civilian non-institutionalized population conducted by NCHS. NHIS collects data from each household member on health conditions, disabilities, doctors’ visits, hospitalization, and other healthrelated topics, as well as demographic and economic data for the household and its members. Most of the core NHIS data was added to the NSFG data file, comprising variables that measure health and demographic characteristics of the respondent, her husband (if married), and her household. The advantages of this linkage included the addition of unusually rich data on women’s health and demographics, the gain of a longitudinal aspect,13 and improved nonresponse adjustments for calculating sample weights. For Cycle 6 (2002), another major change was made in the sample frame: for the first time, men as well as women age 15 to 44 are being interviewed. (Note that these are independent samples, not pairs of men and women.) The questionnaires used for men and women are similar but not identical. The male component can be compared cross-sectionally with the National Surveys of Adolescent Males (NSAM) to examine trends for young men. 14 13 Cycle 4, conducted in 1988, included a re-interview by phone in 1990 which provided a longitudinal dimension. This follow-up had a low response rate, however, and is of questionable validity. NSAM to date comprises four surveys of two groups: a three-wave longitudinal study conducted in 1988, 1990-91, and 1995, and a new cohort survey conducted in 1995. The first cohort (n=1900) was aged 15 to Chapter 6: The National Survey Of Family Growth 49 14 Abt Associates Inc. The NSFG files can be supplemented with a series of “contextual variables,” available to users on special request. These appended measures describe the socioeconomic characteristics of respondents’ communities. Sample sizes: The sample sizes for the first five cycles of the NSFG are shown below. Cycle 1 (1973): 9,797 women aged 15-44 Cycle 2 (1976): 8,611 women aged 15-44 Cycle 3 (1982): 7,969 women aged 15-44 Cycle 4 (1988): 8,450 women aged 15-44 Cycle 5 (1995): 10,847 women aged 15-44 Cycle 6 (2002): 12,000 men and women aged 15-44 (anticipated) Cycle 5 of the NSFG oversampled black and Hispanic women by selecting all households containing black and Hispanic women in the 1993 NHIS. The final sample included 1,553 Hispanic women, 2,446 non-Hispanic black women, 6,483 non-Hispanic white women, and 365 women of other races and origins. Cycle 6 is oversampling blacks, Hispanics, and teenagers. Post-stratified weights are available to make the data nationally representative. Methods of data collection: Throughout its existence, the NSFG has employed in-person interviews in respondents’ homes. CAPI was introduced as of Cycle 5. In addition, Audio Computer-Assisted Self Interviewing (A-CASI) delivered over headphones was used for the section on abortion, to reduce underreporting of this sensitive behavior. (Comparison to other data sources suggested that women reported only about one-third of all abortions in previous NSFG cycles.) Questions regarding some additional sensitive topics, such as forced intercourse, also were asked using A-CASI. Cycle 6 is also using CAPI and A-CASI. 19 in 1988, as was the new cohort in 1995 (n=1700). Both cohorts are nationally representative and oversample blacks and Hispanics. The study is sponsored by the National Institute of Child Health and Human Development. 50 Chapter 6: The National Survey Of Family Growth Abt Associates Inc. Response rates: NSFG reports an overall response rate of 78.6 percent for Cycle 5. This varies very little by race/ethnicity (78 to 80 percent), but somewhat more by respondents’ age, from a low of 75 percent for women aged 25 to 29, to a high of 83 percent for women aged 15 to 17. 15 Survey Content Cycle 5 of the NSFG covered the following topics: respondent’s background (age, marital status, race/ethnicity, household roster, education, employment history, and living arrangements throughout childhood); pregnancy, birth, and adoption history; marital and relationship history; sterilization and contraceptive history; family planning and medical history; birth expectations, desired family size, and infertility services; respondent’s birthplace, religion (affiliation and attendance), and race/ethnicity; self-assessed happiness and well-being; opinions on the roles of men and women; plans to marry current partner; primary job; husband’s/partner’s children and current or last job; earnings, family income and other assistance; and a series of sensitive items administered by A-CASI on abortion, birth control, intercourse partners, forced intercourse, HIV risks, family problems (if under age 20). Cycle 6 covers similar topics for females, but moves work history to near the end of the instrument, excludes the items on self-assessed well-being and family problems, and adds items on health insurance and attitudinal measures on sex, contraception, marriage, and parenthood. For males, the list of topics is as follows: background information as above; sex education, vasectomy, infertility, sexual intercourse, sexual partners; demographic characteristics, fertility, and children of current wife or cohabiting partner; same information for recent sexual partner(s) (up to 3) and first sexual partner; same information for former wives and first cohabiting partner; other biological and adopted children, other pregnancies; fathering (activities with all children, and visitation with and monetary support of non-coresidential children); birth expectations; health conditions, access to health care, receipt of health services; residence, place of birth, rent/own/payment for current residence, religion, military service, work history, work in past year, current or last job (self and partner); child care; attitudes towards sex, contraception, marriage, gender, and parenthood; and a series of sensitive items administered by A-CASI: general health, significant life events (homelessness, incarceration), substance use, pregnancy/abortion, sex with females (including non-voluntary sex and STD/HIV risking behaviors), sex with males (including 15 Response rates by ethnicity: • • • Hispanic: 79.5 percent Non-Hispanic: 77.8 percent Other: 78.8 percent. Response rates by age: • • • • • • 15-17 years: 18-24 years: 25-29 years: 30-34 years: 35-39 years: 40-44 years: 82.5 percent 82.1 percent 74.5 percent 78.0 percent 78.1 percent 78.5 percent. Chapter 6: The National Survey Of Family Growth 51 Abt Associates Inc. non-voluntary sex and STD/HIV risking behaviors), condom use at last sex of any type, sexual orientation and attraction, STDs and HIV, and family income. Suitability for Research Outcome measures: The NSFG provides a good basis for documenting the dynamics of union formation and dissolution. For example, Cycle 5 (Section C) collects information on women’s current and former husbands, cohabitation partners, and other sexual partners as shown below: • Husbands: Date of each marriage, cohabitation prior to the marriage, dates and reasons for separations during the marriage, and (if relevant) when and how the marriage ended. Furthermore, a set of basic demographics are collected for the respondent’s first and last (or current) husband: his date of birth, education, race and ethnicity, religion, the importance of religion to him, and his prior marital status. Current cohabitational partners: When cohabitation began, dates and reasons for separations, and the same demographics listed above. Previous cohabitational partners: Dates of cohabitation. First (male) consensual sexual partner: Abovementioned demographics, marital status then and now, relationship then and now. Other sexual partners: Number in the past year, past four years, lifetime, prior to marriage. Other sexual partners in past four years: Abovementioned demographics, age at first intercourse, relationship then and now, marital status then and now. • • • • • This information supports construction of a history of entrances and exits for sexual relationships, cohabitation, and marriage. Information on union quality is limited: only what can be gleaned from separations, and whether those separations were “because you were not getting along, or for some other reason [e.g., jobs in separate cities, attending school, institutionalized, extended travel], or some of each.” Measures of demographic influences: The NSFG is strong on the measurement of demographic influences. The Cycle 5 instrument and the Cycle 6 instrument for females gather information on all pregnancies including those that did not result in live births, and information on cohabitational as well as marital history. The contraceptive history module collects information on whether each pregnancy was wanted. Related attitudinal measures are also collected, including respondent’s plans to marry. 16 Cycle 6 also collects information on attitudes towards parenthood, from both men and women (see Exhibit 6.5). There are some limitations around the fact that NSFG is a cross-sectional design, so that union histories are reported retrospectively. This makes it impossible to obtain the correct temporal 16 Items include: • • • Do you expect to marry your current boyfriend? Do you expect to marry [again] at some time in the future? At what age do you expect to get married? Would you say you are very sure or not very sure? Chapter 6: The National Survey Of Family Growth Abt Associates Inc. • 52 ordering of attitudes and union behaviors needed for making strong causal conclusions. However, overall the NSFG does a good job of collecting retrospective information on a range of relevant measures. Measures of economic influences: Cycles 5 and 6 have good information about both men and women’s economic well-being. They both ascertain the respondent’s primary job, the spouse’s or partner’s current or last job, earnings, family income, and receipt of public assistance. Income measures are obtained for the last 12 months. Information is also available on the respondent’s work history, and on the educational attainment of the respondent and current and previous partners. The contextual data can provide information about the local labor and marriage markets. The retrospective work history information can potentially be combined with the union histories to better understand the influence of employment factors on union behaviors. Measures of socio-cultural influences: The inclusion of a range of attitudinal measures in the recent waves of the NSFG provides opportunities to examine relationships between attitudes and union behaviors. Attitudinal measures include opinions on “family life and the rights and status of women” (see Exhibit 6.4), items on birth expectations, and desired family size. Unfortunately, the collection of attitudinal data at only one point in time makes it impossible to create the correct temporal ordering of attitudes and union behaviors needed for making strong causal conclusions. Instead, the NSFG can be used to describe cross-sectional differences in attitudes associated with different union status and personal histories, as well as with future marriage plans. Perhaps more importantly, the NSFG offers an excellent resource for examining the relationships between expectations, attitudes and union behaviors indirectly, by linking the union histories with labor force, education, and childrearing information. Additionally, the collection of detailed information on sexual behavior provides opportunities for studying different types of partnerships, and the influence on these partnerships on union behaviors. Measures of psychological influences: The NSFG includes a battery of questions on respondent well-being and anxiety (see Exhibit 6.6). Rich contextual information is available with regard to the respondent’s living situation at birth and throughout childhood; her parents’ marital history, cohabitation, and expectations for respondent’s education; mother’s employment; and the respondent’s current relationship with parents. Respondents under age 20 provide information about their parents’ alcohol and drug abuse and physical violence. Special Strengths: The design of the NSFG offers several special advantages for researchers. First, the ongoing nature of the survey makes it possible to conduct comparisons over two to three decades. Second, the oversampling of blacks and Hispanics in both Cycle 5 and Cycle 6 supports analyses of these subpopulations. Disadvantaged individuals may be identified by household income relative to poverty, receipt of AFDC/TANF, race/ethnicity, education, living arrangements at age 14, and community characteristics. Third, the linkage of Cycle 5 to the NHIS provides a rich source of information about respondents’ and household members’ health. Fourth, Cycle 6 provides similarly detailed information about men as about women. Finally, the availability of contextual data describing respondents’ locales makes it possible to include community characteristics in analyses. The survey also contains data about some types of potential influences that are not often captured in national surveys: the respondent’s living arrangements throughout childhood and as a young adult; the respondent’s views about the roles of men and women; marriage and fertility expectations; and Abt Associates Inc. Chapter 6: The National Survey Of Family Growth 53 current and past partners’ religion and the importance of religion to them, as well as their race/ethnicity. Limitations: Because the NSFG is cross-sectional, dynamic analysis of union formation and dissolution must be performed retrospectively. Characteristics of sample members are measured at the time of the survey, rather than at the time of the events of interest. Although the history of union formation and dissolution is very well documented, there is virtually nothing on union quality. A third limitation of the NSFG is that through Cycle 5, only women were interviewed; information on their male partners comes from the women.17 Also AFDC/TANF participation is captured only in terms of any receipt in the past 12 months. No infor mation is available on current welfare receipt or welfare history. The contextual variables and some of the more sensitive data are not available on the public use files. This information may be obtained from CDC “under special arrangements that assure confidentiality and protection of the data commensurate with that provided by the Center itself.” Details are given below under “Data Availability.” Using the Data The survey website is http://www.cdc.gov/nchs/nsfg.htm. Documentation: The User’s Guide for Cycle 5 is available on-line at the above-mentioned website (National Survey of Family Growth Cycle 5: 1995, User’s Guide, US Department of Health and Human Services, Hyattsville, MD, February 1997). Documentation for previous cycles is provided with the data tapes. Data availability: Cycles 1 through 5 of the NSFG are available on data tape. CD-ROMs are available for Cycle 4 (including the reinterview) and Cycle 5. Data tapes are much more expensive than CD-ROMs. Data collection for Cycle 6 is in the field at this writing and the first reports will be available in early 2004. The NSFG data tapes can be purchased through the National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161, tel. (800) 553-NTIS (6847). Ordering instructions appear on the website: http://www.cdc.gov/nchs/products/elec_prods/intro/ntsorder.htm. Each data tape package contains all the necessary documentation, including the tape contents manual and a description of the sample design, weighting, and variance estimation. CD-ROMs can be obtained directly from NCHS by calling (301) 458-INFO. As noted above, certain data items are omitted from the public use files. Researchers can obtain this information from NCHS under special arrangements by writing to: 17 Given the cross-sectional nature of this survey, the age restriction (15 to 44) is not a significant drawback. Information on older women, which may be of interest in order to compare patterns between earlier and later cohorts, are available from earlier cycles. (In fact, data collected on various cohorts at the same age are preferable to data collected in the same calendar year, because recall bias will not vary systematically across cohorts.) Chapter 6: The National Survey Of Family Growth Abt Associates Inc. 54 National Survey of Family Growth Reproductive Statistics Branch Division of Vital Statistics 6525 Belcrest Road, Room 840 Hyattsville, MD 20782 The restricted items are whether respondent has ever been diagnosed with genital warts, gonorrhea, chlamydia, syphilis, genital herpes; respondent’s current state of residence, classified by DHHS region; day of interview completion (or last attempt); respondent sequence number (ID); all A-CASIobtained data (information on abortions, number of sexual partners, forced sexual intercourse, HIV risk behaviors, and childhood family violence and drinking/drug use); and contextual data linked from the 1990 Census Summary Tape Files, on unemployment rates, poverty rates, etc. Reports and articles: A list of reports and articles using NSFG data is available on their website. Some studies of unions that have used NSFG are: Bumpass, L. & Lu, H. (2000). Trends in cohabitation and implications for children’s family contexts in the United States. Population Studies, 54(1), 29-41. Lichter, D. T. & Graefe, D. R. (2001). Finding a mate? The marital and cohabitation histories of unwed mothers. In L. Wu & B. Wolfe (Eds.), Out of Wedlock: Causes and Consequences of Nonmarital Fertility (pp. 317-343). New York: Russell Sage Foundation. Manning, W. D. (2001). Childbearing in cohabiting unions: racial and ethnic differences. Family Planning Perspectives, 33(5), 217-223. Data Elements The content of the Cycle 5 survey is shown in Exhibit 6.1. The contents of the female and male surveys comprising Cycle 6 are described in Exhibits 6.2 and 6.3, respectively. Attitudinal items from Cycles 5 and 6 are listed in Exhibits 6.4 and 6.5, and the battery of items on well-being and anxiety appears in Exhibit 6.6. Abt Associates Inc. Chapter 6: The National Survey Of Family Growth 55 Exhibit 6.1 Content of NSFG Cycle 5 Instrument Section A: Education, Work, and Family Background Age and marital status, educational attainment and dates of attendance, educational intentions, vocational training and dates of attendance, vocational intentions, suspension and expulsion, lapses in school attendance, general marks received in school, R’s employment history (work history since 18th birthday), living arrangement (information about who R lives with), parents married (parents marital status), living situation at birth and subsequent change(s), relationship status of R’s parents while R was in residence, ever lived with grandparent(s), parents’ relationship, parents’ marital history, family structure, R’s life on own, when, current relationship with parents, presence of formal and/or informal sex education, smoking incidence and intensity. Section B: Pregnancy and Birth History Number of pregnancies, pregnancy outcome and date, delivery method, gestation, costs associated with a nonhospital delivery, presence of medical aid for fertility and prenatal care, smoking incidence and intensity during pregnancy, when learned of pregnancy, child’s living status, breastfeeding, relinquishment for adoption, care of nonbiological children, adoption plans (current and past), adoption preferences. Section C: Marital and Relationship History Number of marriages, first husband and most recent/current husband (date of birth, educational attainment, race/ethnicity, religion, previous marriages, cohabitation before marriage, separations during marriage, reasons for end of marriage), every other husband (date of marriage, cohabitation before marriage, reasons for end of marriage and date, separations during marriage), current partner (date began cohabitation, separations, date of birth, educational attainment, religion, race/ethnicity, previous marriages), other partners (Information on three previous cohabitations), ever had intercourse, first intercourse (date, R’s age, (in)voluntary, date of first voluntary intercourse, R’s age), first intercourse partner (relationship to R, age, marital status, educational attainment, religion, race/ethnicity, R’s current involvement with partner), number of intercourse partners in last 12 months, number of male, sexual partners before marriage, intercourse partners since January 1991 (R’s relationship with partner, age of partner at time of first intercourse, when was first intercourse with partner, when was last intercourse with partner, still considered a sexual partner, marital status of partner). Section D: Sterilizing Operations Sterilizing operations, desire for reversal (for tubal ligations and vasectomies), non-surgical sterility and impaired fecundity. Section E: Contraceptive History Methods used, first method used, periods of non-intercourse, contraceptive method history (method used by month), wantedness and other circumstances surrounding each pregnancy, recent use of birth control, sex and birth control use in past 3 months. Section F: Family Planning and Medical History Birth control services ever received and first received, birth control and medical services in past 12 months, experience with clinics for medical or birth control services. Section G: Birth Expectations and Desired Family Size Wanting a/another baby, intending a/another baby, number intended. Section H: Infertility Services Infertility services, vaginal douching, health problems related to childbearing (including PID and disability), HIV testing. Section I: Background Information R’s birthplace, religion (affiliation and attendance), race/ethnicity, R’s opinion of own happiness and well-being, opinions on the roles of men vs. women, plans to marry current partner, R’s primary job, husband/partner’s children (number of children, living arrangements, child support, husband/ partner’s current/last job, earnings), family income and other assistance. Section J: Audio Computer -Assisted Self-Interview (A-CASI) Number of abortions (date), birth control, recent intercourse partners (sexual partners in last 12 months, since 1991, in lifetime), forced intercourse, HIV risks (R’s own drug use, drug use of partners, any male partners with homosexual histories), family problems - if under age 20 (alcohol/drug abuse amongst parents, physical violence between R’s parents, physical violence between parents and R). Section K: Respondent Locator Information 56 Chapter 6: The National Survey Of Family Growth Abt Associates Inc. Exhibit 6.2 NSFG: Content of Cycle 6 Female Questionnaire Section A: Background, Demographic Information Age, marital/cohabitation status, race/ethnicity, household roster (name, usual residence, gender, age, relationship to R, where is current husband/partner living, husband/partner’s relationship to children), educational attainment, childhood/parental background. Section B: Pregnancy History and Adoption-Related Information Menarche, current pregnancy status, number of pregnancies, detailed pregnancy history, age of father of pregnancy, relinquishment for adoption, care of non-biological children, adoption plans (current and past), adoption preferences. Section C: Marital and Relationship History Marriage and cohabitation history, husband/partner characteristics, first sexual intercourse and characteristics of first partner, reasons for not having sex (among virgins), sex education (teens only), numbers of sexual partners, recent (past 12 months) partner history (R’s relationship with partner, age of partner at time of first intercourse, date of first/last intercourse with partner, still considered a sexual partner, educational attainment, race/ethnicity, current relationship). Section D: Sterilizing Operations and Impaired Fecundity Sterilizing operations, desire for reversal (for tubal ligations and vasectomies), non-surgical sterility and impaired fecundity. Section E: Contraceptive History and Wantedness Methods used, first method used, periods of non-intercourse (last 3 years), method used each month (last 3 years), method used with partners in past 12 months, wantedness and other circumstances surrounding each pregnancy, consistency of condom use, frequency of sex in past 4 weeks. Section F: Family Planning and Medical Services Birth control and medical services in past 12 months, provider and payment information for each visit (more detail if clinic cited) and whether regular source of medical care, first birth control service (date and details), ever visited a clinic. Section G: Birth Desires and Intentions Wanting a/another baby, intending a/another baby, number intended. Section H: Infertility Services and Reproductive Health Infertility services, vaginal douching, health problems related to childbearing (including PID and disability), HIV testing. Section I: Demographic Information and Attitudes Questions Health insurance (health insurance and medical expense coverage over past 12 months), residence, place of birth, rent/own/payment for current residence, religion, work background and in past year, current or last job (R and partner), child care, attitudes towards sex, contraception, marriage, gender, and parenthood. Section J: Audio Computer-Assisted Self-Interview (A-CASI) General health, numbers of pregnancies, substance use, sex with males (including non-voluntary sex and STD/HIV risking behaviors), sex with females, condom use at last sex of any type, sexual orientation and attraction, STD, HIV and risk behaviors, family income (presence of household income from earnings and various public assistance programs). Abt Associates Inc. Chapter 6: The National Survey Of Family Growth 57 Exhibit 6.3 NSFG: Content of Cycle 6 of Male Questionnaire Section A: Background Information Age, marital/cohabitation status, race/ethnicity, household roster (name, usual residence, gender, age, relationship to R, where is current wife/partner living, wife/partner’s relationship to children), education, childhood/parental background, marriage and cohabitation (number of marriages, number of cohabitations). Section B: Sex Education, Vasectomy, Infertility, Sexual Intercourse, Sexual Partners Sex education (teens only), vasectomy and infertility, sexual intercourse, reasons for not having sex (among virgins), number of biological children, number of sexual partners in life and last 12 months, listing of up to three most recent partners. Section C: Current Wife or Cohabiting Partner Cohabitation and marriage, demographic characteristics, first sex with current wife-partner (R’s age, relationship to partner, “safe” sex practices), sterilizing operations and infertility, last sex, summary of method use in last 12 m onths, biological children (number, name, gender, date of birth, marital status to child’s mother, child’s living situation), current pregnancy, her children (number, did R legally adopt any of the children, gender, living situation), other children under their care. Section D: Sexual Partners Cohabitation and marriage, demographic characteristics, first/most recent sexual experience (R’s age, name of partner, age of partner, relationship to partner, “safe” sex practices), summary of method use in last 12 months, biological children (name, gender, date of birth, relationship with partner at time of child’s birth, child’s current living situation), current pregnancy, her children, other children under their care, first sexual partner. Section E: Former Wives and First Cohabiting Partner Cohabitation and marriage, demographic characteristics, biological children (number, name, gender, relationship to partner at time of birth, child’s current living situation), her children (number, did R legally adopt any of the children, gender, living situation), other children under their care. Section F: Other Biological and Adopted Children, Other Pregnancies Additional biological children (number, name, name of child’s mother, gender, date of birth, relationship to mother at time of birth, child’s current living situation), additional adopted children, pregnancies that did not end in live birth. Section G: Fathering Activities with coresidential children, visitation and activities with non-coresidential children, monetary support of non-coresidential children. Section H: Birth Expectations Desires and intentions for a future birth. Section I: Health Conditions, Access to Health Care, Receipt of Health Services Usual source of medical care, health insurance, use of family planning clinic, disability, health services in last 12 months, medical help for infertility, HIV testing. Section J: More Background Information and Attitudes Questions Residence, place of birth, rent/own/payment for current residence, religion, military service, work background and in past year, current or last job (R and partner), child care, attitudes towards sex, contraception, marriage, gender, and parenthood. Section K: Audio Computer-Assisted Self-Interview (A-CASI) General health, significant life events (homelessness, incarceration), substance use (experience with tobacco, alcohol, marijuana, cocaine, crack, and non-prescription drugs using a needle), pregnancy/abortion, sex with females (including non-voluntary sex and STD/HIV risking behaviors), sex with males (including non-voluntary sex and STD/HIV risking behaviors), condom use at last sex of any type, sexual orientation and attraction, STDs and HIV, family income, public assistance (presence of household income from earnings and various public assistance programs). 58 Chapter 6: The National Survey Of Family Growth Abt Associates Inc. Exhibit 6.4 NSFG: Cycle 5 Attitudinal Items • • • • • • • • • • • • • • • • • • A man can make long range plans for his life, but a woman has to take things as they come. A pre-school child is likely to suffer if his mother works. A working mother can establish just as warm and secure a relationship with her children as a mother who does not work. It is much better for everyone involved if the man is the achiever outside the home and the woman takes care of the home and family. A woman should have exactly the same job opportunities as a man. Men should share the work around the house with women such as doing dishes, cleaning, and so forth. A woman should not let bearing and rearing children stand in the way of a career if she wants it. On the job, men should not refuse to work under women. Women are much happier if they stay at home and take care of their children. Young girls are entitled to as much independence as young boys. Men and women should be paid the same money if they do the same work. Sex seems to exist mainly for the man's pleasure. Women should be considered as seriously as men for jobs as executives or politicians or even President. If anything happened to one of the children while the mother was working, she could never forgive herself. A woman's job should be kept for her when she is having a baby. You usually find the happiest families are those with a large number of children. Many of those in women's rights organizations today seem to be unhappy misfits. There should be free child-care centers so that women could take jobs. Abt Associates Inc. Chapter 6: The National Survey Of Family Growth 59 Exhibit 6.5 NSFG: Cycle 6 Attitudinal Items • • • • • • • • • • • • • • • It is better for a person to get married than to go through life being single. Divorce is usually the best solution when a couple can’t seem to work out their marriage problems. Sexual relations between two adults of the same sex [are all right]/[is always wrong]. Any sexual act between two consenting adults is all right. It is all right for unmarried 18 year olds to have sexual relations if they have strong affection for each other. It is all right for unmarried 16 year olds to have sexual relations if they have strong affection for each other. The rewards of being a parent are worth it, despite the cost and the work it takes. It is okay for an unmarried female to have a child Gay or lesbian adults should have the right to adopt a child. A young couple should not live together unless they are married. A working mother can establish just as warm and secure a relationship with her children as a mother who does not work. It is much better for everyone if they man earns the main living and the woman takes care of the home and family. It is more important for a man to spend a lot of time with his family than to be successful at his career. If you [got pregnant]/[got a female pregnant] now, how would you feel? If it turns out that you do not have any children, would that bother you a great deal, some, a little, or not at all? Note: Bracketed words indicate versions for [females]/[males]. 60 Chapter 6: The National Survey Of Family Growth Abt Associates Inc. Exhibit 6.6 NSFG: Cycle 5 Items on Well-Being and Anxiety • • • • • • • • • • • • • • • • Taken altogether, how would you say things are these days? Would you say that you are very happy, fairly happy, neither happy nor unhappy, not very happy, or very unhappy? Have you ever, in your lifetime, had a period lasting six months or longer when most of the time you felt worried and anxious? Has that period ended or is it still going on? [During that period, did]/[Do] you worry about things that were likely to happen or things that were not likely to happen? Did/do you worry about things that are likely to happen or things that are not likely to happen? Did/do you worry a great deal about things that were not really serious, or did you worry about things that were serious? [During that period, did]/[Do] you have different worries on your mind at the same time? When you were/are worried or anxious, were you also restless? When you were/are worried or anxious, were you also keyed up or on edge? Were/are you particularly irritable? Were/are you aware of your heart pounding or racing? Were/are you easily tired? Did/do you also have trouble falling asleep or staying asleep? Did/do you feel faint or unreal? When did this period end? How many months or years has this period been going on? Have you had any other periods of this sort in your life, that happened before the one you just described? What is the earliest age you can clearly remember having a period of this sort? Abt Associates Inc. Chapter 6: The National Survey Of Family Growth 61 Chapter Seven: A National Survey Of Families And Households The National Survey of Families and Households (NSFH) is a longitudinal survey funded by the Center for Population Research of the National Institute of Child Health and Human Development (NICHHD) and the National Institute on Aging (NIA). Principal investigators are Larry Bumpass and James Sweet, University of Wisconsin, Madison. The focus of the study is the “causes and consequences of changing family and household structures,” and it comprises an unusually rich source of information in this area. Survey Design NSFH is currently in its third wave of data collection from a nationally representative sample of the non-institutionalized population of the United States, aged 19 and older. Wave 1 was collected in 1987-88, Wave 2 in 1992-94, and Wave 3 in 2001-02. NSFH is unique in identifying and interviewing a circle of individuals related to the primary respondent in various ways. In Wave 1, a primary respondent was randomly chosen in each household from the roster of adults. This individual provided information about him or herself and the household in general. Sensitive information was collected via a self-administered questionnaire (SAQ). The primary respondent’s spouse or cohabiting partner, if any, also completed an SAQ as a secondary respondent. Finally, if neither of these two persons was the householder, the householder completed an SAQ as a tertiary respondent. In Wave 2, personal interviews were conducted with the primary respondent, his/her current spouse or cohabiting partner, and other secondary respondents from Wave 1—that is, former spouses and cohabiting partners. In addition, telephone interviews were conducted with primary respondents’ focal children aged 5 to 23, and with primary respondents’ parents (one randomly selected per respondent). Proxy interviews provided information on respondents who had died or were too ill to be interviewed. In Wave 3, telephone interviews were attempted for primary respondents with focal children in Wave 2; primary respondents age 45 or older without focal children in Wave 2; and the Wave 1 spouses and partners of these individuals. Thus, younger childless respondents and their spouses and partners were not followed in Wave 3. Wave 2 focal children were also re-interviewed. Sample size: The survey oversamples blacks, Puerto Ricans and Mexican Americans in order to provide adequate sample sizes for analyses of these racial and ethnic subpopulations. NSFH also stratified their sample based on family formation status oversampling individuals in single -parent families, families with stepchildren, cohabiting couples, and recently married persons. The main sample comprised some 9,600 households and the oversample an additional 3,400 households. Methods of data collection: For primary and secondary respondents, Waves 1 and 2 were collected in person, via CAPI. The SAQs were paper-and-pencil instruments. Wave 3 is being collected from primary respondents by telephone, via CATI. Abt Associates Inc. Chapter 7: A National Survey Of Families And Households 63 Response rates: Response rates in Wave 1 were 74 percent for primary respondents, 83 percent for secondary respondents who were spouses, 77 percent for secondary respondents who were cohabiting partners, and 78 percent for tertiary respondents. Survey Content: Wave 1 Unusually detailed life history data were collected in Wave 1. The CAPI questionnaire covered the following areas: current household composition, respondent’s household history, marriage and cohabitation history, social background of respondent’s first spouse, married respondents living apart from spouse for reasons other than marital conflict, fertility history and expectations, quality of relationships with children, and social and economic characteristics. The SAQ for the primary respondent covered additional topics: household tasks, physical/mental health and social support networks, feelings and plans of respondents living in parental households, divorce and separation experiences, attitudes and considerations of unmarried, non-cohabiting respondents up to age 35, cohabitation relationships, married couple relationships, fertility considerations, parenting, feelings about adult children in household, relationship with children, and information and attitudes about parents, siblings, and in-laws. The secondary respondent (spouse or cohabiting partner of the primary respondent, if any) was administered an SAQ which covered the following areas: birth date, ages lived with natural or stepparents, information about parents, siblings, and in-laws, marriages, cohabitation, divorce and separation, births, fertility intentions and considerations, relationships with children, information on absent parent of child currently in household, information on children up to age 18 not living in household, information on adult children living in household, military experience, education and occupation, household tasks, occupation at age 50, and attitudes on topics such as mothers who work, fertility, employment, sex roles, self-esteem, non-marital and extra-marital sex, kinship help, and religiosity. Topics addressed in the SAQ administered to the tertiary respondent (i.e., the householder, if not previously questioned) were as follows: householder’s relationship with adult child in household, householder’s relationship with other adult relative living with householder, household tasks, marital history, quality of relationship, areas of disagreement with spouse, relationship with children, alcohol or drug problems, information about householder and spouse, household income, assets and debts, and attitudes on the same topics listed above. Survey Content: Wave 2 Wave 2 included four types of interviews: the main interview, the parent interview, the younger focal child interview, and the older focal child interview. The content of each of these is summarized below. The main interview was administered to the NSFH1 primary respondent, the NSFH1 spouse/partner, and (if different), the current spouse/partner of the NSFH1 primary respondent. Topics covered included: current household composition, children elsewhere, goings and comings of children, caregiving and receiving, co-residence with parents, relationships with adult children and siblings, grandparenting, characteristics of new spouses (filling in information that had been collected at Wave 64 Chapter 7: A National Survey Of Families And Households Abt Associates Inc. 1 for the NSFH1 primary respondent and NSFH1 spouse/partner), marriage, cohabitation, dating, widowhood, fertility, problem inventory for all children aged 5-17, behavior and outcomes for focal child aged 5-17, problem inventory for all children aged 18-23, relationship with focal child aged 1823, problem inventory for children under age 5, new focal child under age 5, resident child’s absent parent, absent child’s other parent, residence history, religion, education history, work history, current employment, income, assets, and debt. Self -administered modules collected data on household tasks, health and well-being, recent marital disruptions, attitudes and expectations of unmarried non-cohabiting respondents, attitudes, expectations, and satisfaction of cohabiting and married respondents, behavior problems, moods, and parental expectations for a focal child aged 5-17, behavior problems and moods for a focal child aged 3-4, parenting, child development, family attitudes, social participation, and work. The second type of interview was administered to the NFSH1 respondent’s randomly selected parent. This interview covered household composition, marital history, health and well-being, caregiving and receiving, relationships with parents, relationships with adult children, siblings, grandparenting, current employment, income, assets, and debts. A third type of interview was administered to focal children aged 10 to 17. This covered schooling, friends, psychological well-being, activities, attitudes, parents and siblings, relationship with absent parent, relationship with biological parent (for children who lived with neither parent), and relationship with grandparents (we do not tabulate elements in this interview). The final interview type was administered to focal children aged 18 to 23. Topics covered were: living arrangements and history, marital history, education, dating, fertility, relations with mother, age, religion, employment and income, relationships with father, partner relationships, social support, attitudes and psychological well-being, sex and drugs, and relations with stepparents, siblings, and grandparents. Survey Content: Wave 3 The Wave 3 interview of main respondents and their spouses was similar in coverage to the Wave 2 main interview. Focal children (now ages 15-33) were interviewed on the subjects of household composition, separations from biological parents, leaving parent’s home, marital history, cohabitation history, high school education, post-secondary education, dating, fertility, relationship with mother, employment experience, income, household tasks, relationship with father, happiness with marital relationship, happiness with cohabiting relationship, social integration and social support, religion, family attitudes, health and well-being, sexual activity, use of tobacco, alcohol, and marijuana, problem drinkers in family of origin, and relationship with step-parent, siblings, and grandparents. Suitability for Research Outcome measures: The NSFH offers excellent measures of union formation, quality and dissolution, with generally parallel measures available about marriage and cohabitation. The NSFH provides information on union status at each interview and a full history of marital and cohabitational Abt Associates Inc. Chapter 7: A National Survey Of Families And Households 65 relationships. Rich data on union quality are available as well, from both partners, including time spent on household tasks, experiences surrounding divorce and separation, perceived union quality, areas of disagreement, ways of coping with disagreement, and physical arguments. Measures of demographic influences: The NSFH is very strong in its measure of demographic influences, as well as its measures of related attitudes. Its collection of complete marriage, cohabitation, and fertility histories make it an obvious choice for studying the effects of teen and nonmarital childbearing, transitions to parenthood, and cohabitation on marriage. Respondents’ expectations for marriage, cohabitation, and fertility are also measured. Additionally, the collection of union and fertility histories from focal children in the third wave should permit investigation of the influence of early childbearing on union behaviors in a recent cohort. This new data can also be compared to the experiences of the original adult respondents, permitting tests of cohort or period influences. Measures of economic influences: The NSFH collects more than adequate information about relevant economic influences. Information is collected on the respondent’s education, employment history, current employment (occupation, hours, earnings), work schedule, and commute. In contrast to other surveys which rely on proxy-reporting, in most cases these measures are available directly from each partner in married or cohabiting households. Annual household income information includes earnings for all household members and receipt and amounts of public assistance, Social Security, and other forms of unearned income. Intergenerational economic transfers between households are also documented. The NSFH public use files formerly included several useful site characteristics (urban/rural, poverty, race, income, education levels, industry, metropolitan status), but no longer do so. The NSFH website instructs users who wish to merge site-level data with individual records to contact NSFHHELLP@sss.wisc.edu. This presumably gives researchers the ability to include information on labor markets, marriage markets, and public policies in their analyses. Measures of socio-cultural influences: NSFH provides unusual breadth and depth in its collection of attitudinal measures likely relevant to union behaviors. For cohabiting respondents, these include the relative importance of reasons for and against cohabitation (e.g. “it requires less personal commitment than marriage,” “my parents disapprove”), how life would be different if one were married (e.g. standard of living, overall happiness, sex life), relative importance of marital timing issues (“having enough money saved,” “being established in your job”), willingness to marry someone with various traits (“already had children,” “was not ‘good looking’”), and general attitudes towards marriage. Non-married, non-cohabiting respondents are similarly asked about marital timing considerations, marital considerations, and attitudes for and against cohabitation. No information is collected about gender roles. Many of the same attitudinal measures were measured repeatedly in each wave, permitting analysis of changes in attitudes, and linking these changes to union behaviors. Since similar attitudinal items were asked of the original sample members and their adult focal children in Wave 3, there should be an opportunity to investigate cross-sectional trends in attitudes about marriage and cohabitation, as well as to study the impact of parents’ attitudes on children’s union behaviors. Measures of psychological influences: In addition to the multiple measures of union quality discussed above, the NSFH collects measures of a range of relevant psychological influences. The self-administered questionnaire addresses a variety of psychological domains: happiness, depression, 66 Chapter 7: A National Survey Of Families And Households Abt Associates Inc. evaluation of roles,18 perceived health, alcohol and drug problems, physical and mental limitations, and social activities. There are also measures of different aspects of interpersonal relationships with children, parents and other household members. Measures of psychological influences at Wave 1 or Wave 2 can be combined with information in union quality or changes in union status measured at a later wave of the survey to develop causal models with correct temporal ordering. The NSFH also collects background information on respondents’ childhood household history: siblings, leaving home, and periods when the respondents’ parents lived with him or her. Among measures of contextual influences, items on parents’ occupation and family receipt of public assistance during the respondent’s youth are also included, as are items on religious preference and activities. For the cohort of focal children, information about their problem behaviors and relationships with their parents during childhood and adolescence is also available. Special strengths: The NSFH is an exceptionally strong resource for analyzing the determinants of union formation, quality, and dissolution. Favorable design features include oversampling of key groups, detailed retrospective data on marital and cohabitational history, multiple waves of interviews, collection of information from primary respondents’ spouses and cohabitational partners, and rich descriptive information on many potential influences, including feelings and attitudes, social background, psychological well-being, and others. Another strength of the NSFH is its potential to examine union behaviors among the disadvantaged, due to the oversampling of blacks, Puerto Ricans and Mexican Americans. The separate oversampling of the latter two Latino groups permits more detailed consideration of the diversity of cultural experiences and behaviors than possible in many other data sources. Additionally, measures of educational attainment, income, and family background are adequate for developing alternate approaches to the definition of disadvantage, and examining associated variations in union behaviors. The design of the third wave of data from the NSFH suggests additional strengths in two areas. First, it extends the longitudinal nature of the data, providing further information about individuals life course trajectories. Second, the new data should permit inter-cohort comparisons of the experiences of the original respondents and their focal children, to consider changes over time in union behaviors and their influences. Wave 3 is not yet publicly available, and the value of the data for both of these issues will depend in part of careful analyses of issues of attrition and representativeness. Limitations: While no survey is without relative weaknesses, the NSFH is tailor-made for a broad research agenda on union behaviors, especially in the socio-cultural domain. The broad spacing of the observations (seven or eight years between waves) renders these data inappropriate for analyses that focus on short-run changes. Using the Data Information on the NSFH is available on the NSFH website, at http://www.ssc.wisc.edu/nsfh/home. htm. 18 For each of four roles—“the work you do around the house,” “the work you at your paid job,” “the things you do as a parent,” “the things you do as a husband or wife,” the respondent indicates on a scale of 1 to 7 whether they are interesting versus boring, appreciated versus unappreciated, overwhelming versus unmanageable, complicated versus simple, lonely versus sociable, and poorly done versus well done. Chapter 7: A National Survey Of Families And Households 67 Abt Associates Inc. Documentation: Technical documentation is available at ftp://elaine.ssc.wisc.edu/pub/nsfh. Key reports are: Sweet, J.A., Bumpass, L.L., & Call, V. (1988). The design and content of the National Survey of Families and Households. Center for Demography and Ecology, University of Wisconsin-Madison, NSFH Working Paper No. 1. Sweet, J.A. & Bumpass, L.L. (1996). The National Survey of Families and Households Waves 1 and 2: Data Description and Documentation. Center for Demography and Ecology, University of Wisconsin-Madison (http://www.ssc.wisc.edu/nsfh/home.htm). Wave 1 and Wave 2 NSFH data can be downloaded free of charge from the survey website. Instructions are provided at http://www.ssc.wisc.edu/nsfh/avail.htm. Reports and articles: A list of publications based on the NSFH data can be found at http://www.ssc.wisc.edu/nsfh/bib.htm. Three examples of studies of union formation and dissolution that have used the NSFH are: Koball, H. (1998). Have African American men become less committed to marriage? Explaining the twentieth century racial cross-over in men’s marriage timing. Demography, 35(2), 251-258 Sayer, L. C. & Bianchi, S. M. (2000). Women’s economic independence and the probability of divorce: A review and reexamination. Journal of Family Issues, 21(7), 906-943. Smock, P. J. & Manning, W. D. (1997). Cohabiting partners’ economic circumstances and marriage. Demography, 34(3), 331-341. Data Elements The contents of the NSFH are described in the tables that follow. I. Wave 1 A. Interview with main respondent: Exhibit 7.1 B. Self-administered questionnaire (SAQ) for main respondent: Exhibit 7.2 C. Other components of Wave 1: SAQ for respondent’s spouse, SAQ for respondent’s resident partner, SAQ for tertiary respondent: Exhibit 7.3 Wave 2 A. Main interview: primary respondent, current spouse/partner, Wave 1 spouse/partner: Exhibit 7.4 B. SAQ: Exhibit 7.5 C. Parent questionnaire: Exhibit 7.6 D. Older focal child questionnaire (age 18-23): Exhibit 7.7 Wave 3 Chapter 7: A National Survey Of Families And Households Abt Associates Inc. II. III. 68 A. Main interview: primary respondent, spouse: Exhibit 7.8 B. Focal children questionnaire: Exhibit 7.9 IV. Measures of Union Quality and Attitudes: Exhibit 7.10 Abt Associates Inc. Chapter 7: A National Survey Of Families And Households 69 Exhibit 7.1 NSFH – Wave 1: Content of Interview with Primary Respondent Household Composition Detailed household composition, children living in household, disability or chronic illness; care and assistance given and received. Household History Childhood household history, step-, half-, and full siblings, home-leaving experience, periods when R’s parents lived with R. Marriage and Cohabitation History Marriage history, cohabitation history (currently or previously married), cohabitation history (never married), dating activities of not currently married; marriage and cohabitation plans. Social Background of First Spouse Previous marriages and children, education, socioeconomic background, religion, and intact family, marital separations during first marriage. Married, Spouse Absent from Household Reasons, frequency of contact, expected duration. Fertility History and Expectations Birth history, fertility intentions. All Children Age 5-8 Behavior problems, difficulty (or ease) in raising. Children Age 0-18 Select one focal child. R’s educational expectations for child, separations from child, R’s and spouse/partner’s difficulty dealing with child. Children With Only One Parent in Household Select one focal child. Information on absent parent regarding residence, marital status, other children, child's contact with other parent, R’s contacts with child's other parent, conflict over child, legal agreements regarding residence, visits, child support, alimony, child support payments. Step-Child/Partner’s Child Select one focal child. Residence of absent parent of spouse's/partner’s child, child's contact with other parent, child support payments. Children with No Biological Parent in Household Select one focal child. Residence of child's absent parents, child's contact with parents, child support payments. Biological Children Under Age 19 Living Elsewhere Select one focal child. When last lived with R, current residence, information on other parent regarding residence, marital status, other children, R’s contact with child, conflict over child, legal agreements regarding residence, visits, child support, alimony, child support payments. Spouse/Partner’s Biological Children Under Age 19 Living Elsewhere Select one focal child. When last lived with R, current residence, spouse's/partner’s contact with child, child support payments. 70 Chapter 7: A National Survey Of Families And Households Abt Associates Inc. Exhibit 7.1 NSFH – Wave 1: Content of Interview with Primary Respondent Children/Step-Children Age 19 or Older in Household OR Children/Step-Children of Any Age Away from Home Attending College Select one focal child. School attendance, financial aid, R’s contribution toward schooling expenses, child's living arrangements, child's employment and earnings, child's payments to R for room or board, R’s payments of child's expenses. All Biological and Stepchildren Over Age 19 not Away at School, but Living Elsewhere Children's age, marital status, own children, current residence, R’s contact with children. Social Background Race, religious preference and activity, recent residential movement, parent's occupation and education, family's receipt of public assistance during R’s youth. Secondary and Postsecondary Educational History High school or GED diploma, postsecondary enrollment history, degrees and certificates. Military Service Employment Employment history, current occupation, hours, earnings, current second job, current work schedule, travel to work and away from home, childcare arrangements while at work, work experience in 1986, occupation at age 50. Income, Assets and Debts Income, economic relationships with other relatives living in the household, r living in household of parent or relative, first home purchase, inter-household economic transfers, assets and debts, interviewer observations. Abt Associates Inc. Chapter 7: A National Survey Of Families And Households 71 Exhibit 7.2 NSFH – Wave 1: Content of Self-Administered Questionnaire: Primary Respondent SE-1: Household Tasks For all respondents. Hours per week that the respondent, the respondent's spouse, and others in the household spend doing 9 different types of household tasks. SE-2: Well-Being, Role Performance, Health, Social Participation, Social Support For all respondents. Happiness (global), depression scale (12 items), evaluation of roles, health (global), alcohol or drug problems, physical or mental limitations, sources of help outside household, social activities, participation in organizations (15 types), help provided to and received from people outside household, help provided to and received from each adult child. SE-3: Feelings and Plans Regarding Living with Parents For respondent living in parental household. Living arrangement (global), amount paid to parents, future residence plans, areas of disagreement with parents (10 areas), perceived alteration of life if R moved. SE-4: Divorce and Separation Experience For respondents divorced or separated after January 1, 1977. Divorce/separation instigator, current relationship with former spouse (global), sources of emotional support during separation (8 sources), spouse's income level prior to separation, verbal and physical arguments prior to separation, resumption of dating, R’s and spouse's involvements with others prior to divorce, contact with former spouse, perceived changes in life due to divorce (9 areas). SE-5: Feelings about Marriage and Cohabitation For non-married, non-cohabiting respondents age 35 and younger. Perceived changes in life if R married (9 areas), marital timing considerations (5 items), marriage considerations (12 items), sexual frequency, attitudes for and against cohabitation (13 items), attitudes toward. SE-6: Cohabitation Relationships For cohabiting respondents. Marriage plans, quality of relationship (global), fairness in role allocation, time spent together (global), sexual frequency, areas of disagreement with partner (7 areas), coping with disagreement, physical arguments, perceived changes in life if separated (6 areas), probability of separation, attitudes for and against cohabitation (13 items), perceived changes in life if R married (9 areas), marital timing considerations (5 items), marriage considerations (12 items), attitudes toward. SE-7: Married Couple Relationships For married respondents. Quality of relationship (global), type of marriage ceremony, fairness in role allocation, time spent together (global), sexual frequency, areas of disagreement with partner (7 areas), coping with disagreement, physical arguments, perceived changes in life if separated (6 areas), probability of separation SE-8: Fertility Considerations For respondent that is female age 39 or younger, a single male age 44 or younger, or a married male whose wife/partner is age 39 or younger. Things that people consider when thinking about having a child or another child. (16 items) SE-9: Parenting For respondents where all children are under age 5. Time spent with children, discipline, R’s desires for child's behavior (12 items), step-parenting (8 items). 72 Chapter 7: A National Survey Of Families And Households Abt Associates Inc. Exhibit 7.2 NSFH – Wave 1: Content of Self-Administered Questionnaire: Primary Respondent SE-10: Parenting For respondents with one or more children age 5-18. Meals with children, time spent with children, discipline, R’s involvement in youth groups, R’s desires for child's behavior (12 items), step-parenting (8 items). SE-11: Feelings About Adult Child Living in Household For respondents with adult son or daughter living in household. Living arrangement (global), meals with child, time spent with child, perceived alteration of life if child moved, areas of disagreement with parents (10 areas), frequency of enjoyable or difficult times with child, R’s perception of child's future residence plans, coping with disagreement, physical arguments. SE-12: Relationship with Children For respondents who have, or spouse/partner has, living children. Quality of relationship with each child (global). SE-13: Parents, Relatives, and General Attitudes For all respondents. Information about mother, information about father, information about step-parent, brothers and sisters, inlaws, attitudes. Abt Associates Inc. Chapter 7: A National Survey Of Families And Households 73 Exhibit 7.3 NSFH – Wave 1: Content of Other Sections Self-Administered Questionnaire: Husband or Wife of Primary Respondent Birth date, ages lived with natural or step-parents, information about mother, information about father, brothers and sisters, in-laws, marriages and births, characteristics at marriage, cohabitation, current marriage, fertility intentions, fertility considerations, relationships with children, information on absent parent of child currently in household (select focal child), information on child age 18 or younger who does not live in household (select focal child), adult child living in household (select focal child), divorce and separation (If R was divorced or separated after January 1, 1977), information about R, military, education, and occupation, household tasks, income, occupation at age 50, attitudes. Self-Administered Questionnaire: Cohabiting Partner of Primary Respondent The cohabiting partner of the primary respondent completes a questionnaire, most of which is identical to that completed by spouses, except for minor changes in wording. In the husband or wife questionnaire, questions 50 - 68 deal with marriage and cohabitation prior to marriage. This section was replaced in the partner questionnaire with the following: Dates of birth for all children (C50-C52P08F) Times married (C53-C58) (if previously married) How prior marriage ended (C59-C61) Information on current cohabitation relationship (C62M-C67) 1. Date cohabitation started 2. Marriage plans In the main questionnaire married respondents were asked the income, assets, and debt, of themselves and their spouses. Questions on these topics were not included in the spouse questionnaire. Cohabiting primary respondents were asked only about their own economic situation, and not about the situation of their partner or other household members. To complete the economic picture, a section on the partner’s economic situation was included in the partner’s self-administered questionnaire. Tertiary Questionnaire A self-administered questionnaire is given to the householder when the primary respondent is the son, daughter or other relative of the householder. Householder’s relationship with adult child living in household, householder’s relationship with other adult relative living with householder, household tasks, marital history, quality of relationship (global), areas of disagreement with spouse (6 areas), relationship with children, alcohol or drug problems, information about householder and spouse, household Income, assets and debts, attitudes. 74 Chapter 7: A National Survey Of Families And Households Abt Associates Inc. Exhibit 7.4 NSFH – Wave 2: Content of Main Interview: Primary Respondent, Current Spouse/Partner, ex-Spouse/Partner Household Composition Birth date, marital status, detailed household composition, householder, if absent spouse: reason for absence. Sons and Daughters Living Elsewhere Own and spouse/partner’s. Goings and Comings of Children Children on roster at T1 and not mentioned in above, children on roster at both T1 and T2, children on roster at T2 but not T1, children on roster at T1 but not T2, children under 18 on roster at neither T1 nor T2. Caregiving and Receiving Long-term physical or mental conditions for all household members, anyone in household need assistance, R given/received assistance in last 12 months to/from anyone not living with them, R given/received help in last 12 months to/from anyone living with them. Relationships with Parents Mother, father, spouse/partner’s mother, spouse/partner’s father, step-parents: living, age, date died, health (not for step), relationship with, marital status, where live, frequency of contact, parent types alive at T1: ever live in R’s household since T1, own parents: ever live in household of R’s siblings since T1; ever patient in nursing home or long-term care facility since T1, all parent types: help given and received in past month, hospitalization, physical and mental health, income problems. Coresidence with Parents R lived with parent both T1 and T2: anytime not living with parent since T1, R lived with parent T1 not T2: date and reason apart, R lives with parent T2 not T1: date and why began living with parent, R lived with parent at neither T1 nor T2: live with parents at any time since T1. Relationships with Adult Children All children (and spouse/partner’s children) aged 19 or older, help given/received in past month. Siblings Number of living siblings including step and half, how far away, how often see any, phone or letter. Grandparenting For all parents with children who have children. Number of grandchildren, age youngest, age oldest, nights which any grandchild stayed overnight without parent in past 12 months, frequency contact in last 12 months, relationship quality, responsibility for a grandchild for 6 months or more. Characteristics of New Spouses Parental family intact, mother’s education, father’s education, race/ethnicity, T1 characteristics. Marriage, Cohabitation, Dating and Widowhood Dates of marriages, separations, divorces and widowhoods since T1, dates of beginning and end of all cohabitations since T1, characteristics of partners for those married since T1 (or first cohabitation partner if no marriage since T1, if neither cohabiting nor married at T2: how likely to ever marry, cohabit; dating; if have steady partner, widowed since T1: cause of death, medical conditions, care required, where lived, help from others, when last worked, characteristics if spouse was a nonrespondent at T1. Fertility For women under age 50 and men with a wife/partner under age 50. Biological and adopted children since T1, currently pregnant, future birth plans. Abt Associates Inc. Chapter 7: A National Survey Of Families And Households 75 Exhibit 7.4 NSFH – Wave 2: Content of Main Interview: Primary Respondent, Current Spouse/Partner, ex-Spouse/Partner Problem Inventory, all Children 5-17 Difficulty or ease in raising, long-lasting physical condition, mental or emotional problem, dropped out of school or not been able to attend school for at least 3 months, repeated a grade, in trouble with police, skipped school or cut classes, gotten [someone] pregnant while not married, seen a therapist for emotional problems since T1. Focal Child Age 5-17, Behavior and Outcomes Global health, height, weight, grade in school, average grades, good at athletics or sports, art or design, music, making or fixing mechanical things, total education parent thinks child will get, friends, spend time last week in one-on-one activities, rewards and punishments, frequency in last 30 days talk about: something worrying child, something child was excited about, given child a hug/kiss or spanked child in past week, how influence, allowed to be home alone at certain times, curfew, r and child's contact about school, school child attends, R’s involvement in child's school, TV watching, chores and allowance, other money earned, proportion of child's earnings saved and spent, driver’s license, vehicle ownership and access, dating, open disagreements, overall quality of relationship, child's relationship with R’s parents. Problem Inventory, Children Age 18-23 Difficulty or ease in raising, long-lasting physical condition, mental or emotional problem, dropped out of school before completing high school, in trouble with police, gotten [someone] pregnant while not married, seen a therapist for emotional problems since T1. Relationships with Focal Child Age 18-23 Time spent together, open disagreements, avoid talking about certain topics because leads to arguments, global relationship, how handle disagreements, relationship quality, happiness with child's schooling, relationships and occupation, likelihood of talking with child, how influence, if child living here: how works out, pay for room or board, amount, how much R pays for child's expenses. Problem Inventory, Children Under Age 5 Difficulty or ease in raising, long-lasting physical condition, mental or emotional problem, seen a therapist for emotional problems since T1. New Focal Child Under Age 5 Global health, height, weight, ready for kindergarten, behaviors during feeding, sleeping and eating patterns, crying and fussing, one-on-one time spent, TV watching, preschool, time spent per day taking care of child, relationship, child's relationship with R’s parents Resident Child’s Absent Parent Children before those with R, characteristics at time child was born, current marital/cohabitation status, had children since, miles away, work and income, contact with child, how often would like child to spend time with, grandparents on other parent's side, influence other parent has, frequency talk with about child, conflict over child, R’s satisfaction with other parent, amount of child support, legal agreement about custody and visitation. Absent Child’s Other Parent Children before those with R, current marital/cohabitation status, had children since, state, work and income, contact with child, how often would like to spend time with child, grandparents on R’s side, influence R has, frequency talk about child, conflict over child, R’s satisfaction with situation and other parent, amount of child support paid, legal agreement about custody and visitation. Residence History Since T1, any other address or city; for each intercity move, date, city, state, and number of addresses at that location. Religion Religious preference, denomination, frequency attend. 76 Chapter 7: A National Survey Of Families And Households Abt Associates Inc. Exhibit 7.4 NSFH – Wave 2: Content of Main Interview: Primary Respondent, Current Spouse/Partner, ex-Spouse/Partner Education History If less than high school grad at T1: since T1 high school diploma or GED, dates of each spell of enrollment in each school type since T1 (mostly full-time or mostly part-time), degrees since T1 and dates, active duty in the military since T1. Work History Whether still at T1 job, dates of each spell of employment and full or part-time, job searching when not working, if working continuously since T1: hours worked. Current Employment Main job: hours worked, employer, occupation, wage, work at home; schedule: days and times, minutes travel to work; second job: hours, seasonal or year round, hours work at home; schedule: days and times, how many of past 52 weeks: working, unemployed, out of labor force, looked for work in last 4 weeks, last job: wage rate, occupation, same occupation as at age 50 (if aged 60-64), spouse's work. Incomes, Assets, and Debts Amount of income for R, and if householder or spouse of householder, for each household member age 16+, years since 1987 in which received public assistance, if R is householder and has adult parent or child in household, payments received, living expenses R pays, gifts or loans between R and person, if R is not householder and not a child of householder: payments made, living expenses R pays, gifts or loans between R and person, home purchases, inheritances, gifts or loans from/to friends or relatives, help given/received for living or educational expenses, value of home, real estate and vehicles, amount of savings and investments, debt amounts and payments, monthly rent. Interviewer Observations, Contacting Focal Child and Parent for Interviews, Interview Date Residence structure, who present during interview, respondent's: ability to understand, cooperativeness, interest, amount of interruption, rapport, contact information for focal child and parent interviews, contact person for R, date of interview. Abt Associates Inc. Chapter 7: A National Survey Of Families And Households 77 Exhibit 7.5 NSFH – Wave 2: Content of Self-Administered Questionnaire SE-1: Household Tasks For all respondents. Hours per week R, spouse, others spend doing household tasks. SE-2: Health and Well-Being For all respondents. Global happiness, satisfaction, global health, physical and mental conditions which limit activities, partial depression scale, self-esteem, efficacy, alcohol use. SE-3: Recent Marital Disruptions For respondents with marital disruption since T1. Who wanted marriage to end, description of marriage, arguing, dating, children's feeling about dating and remarriage, involvement with others before end of marriage, communication with former spouse, spouse's parents, friends, where lived, income before separation, state divorced. SE-4: Unmarried, Non-Cohabiting For respondents neither married nor cohabiting. How life would be different if married, steady relationship, frequency of sex, attitudes, desired number of children and employment hours. SE-5: Cohabitation For cohabiting respondents . Global relationship, future marriage plans, happiness, fairness in chores, work, money and child care, time together, frequency sex, desired number of children and employment hours, disagreements, differences in life if were to separate, stability of relationship, attitudes. SE-6: Marriage For married respondents. Global relationship, future marriage plans, happiness, fairness in chores, work, money and child care, time together, frequency sex, desired number of children and employment hours, disagreements, differences in life if were to separate, stability of relationship. SE-7: Focal Children Aged 5-17 For parents of focal children aged 5-17. Behavior problems and moods inventory (27 items from NLSY), importance of goals for children: high school, college, job, career, money, marry, children, leader, live nearby. SE-8: Focal Chil dren Aged 0-4 For parents with focal children aged 0-4. Behavior problems and moods inventory (27 items from NLSY). SE-9: Parenting For parents of children of all ages. Frequency spend time with children, frequency discuss important decisions with children, frequency change rule because of child's opinion, family descriptors. SE-10 to SE-17: Development, Ages 0-4 For parents of children under age 5. Fifteen developmental measures from NLSY, targeted to ages in months: 0- 3, 4-6, 7-9, 10-12, 13-15, 16-18, 19-21, and 22-47. SE-18: Family Attitudes, Social Participation, and Work For all respondents. Gender roles, cohabitation, marriage, marital stability, family obligations, parenthood, unwed parenthood, retirement, social participation, help given and received, health insurance and pension coverage, self -definition concerning retirement (ages 55+). 78 Chapter 7: A National Survey Of Families And Households Abt Associates Inc. Exhibit 7.6 NSFH – Wave 2: Content of Parent Questionnaire Household Composition Household composition, R’s marital status, age, education, children and grandchildren, if absent spouse, reason for absence, householder, contact with children, children living elsewhere, global quality of relationship. Marriage Times married, for each: when began, how end, date death, divorce, separation; if cohabiting, date began; if unmarried: how likely will get (re)married someday, times dated in last month. Health and Well-Being Global happiness, frequency worry about family's expenses and bills, global health, marital happiness, physical or mental conditions that limit activities, depression scale, alcohol use, attitudes on obligations between parents and children, social participation, church attendance. Caregiving and Receiving Long-term physical or mental conditions for all household members, anyone in household need assistance, R given/received assistance in last 12 months to/from anyone not living with them, R given/received help in last 12 months to/from anyone living with them. Relationships with Parents Mother, father, stepparent, spouse/partner’s mother, spouse/partner’s father: living, age, date died, health (not for step), relationship with, marital status, where live, frequency of contact, parent types alive at T1: ever live in R’s household since T1; ever patient in nursing home or long-term care facility since T1, own parents: ever live in household of R’s siblings since T1, all parent types: help given/received in past month, hospitalization, physical and mental health, income problems. Relationships with Adult Children All children (and spouse/partner’s children) aged 19 or older, help and support given/received in past month, all spouse/partner’s children: long lasting physical condition; long lasting mental or emotional problem. Siblings Number of siblings including step and half, distance siblings live, frequency contact. Grandparenting For all respondents whose children have children. Number of grandchildren, age youngest, age oldest, nights which any grandchild stayed overnight without parent in past 12 months, frequency contact in last 12 months, relationship quality, responsibility for a grandchild for 6 months or more. Current Employment Self and spouse: employed, (if not) when last worked, usual hours, occupation; how many of past 52 weeks: working, unemployed, out of labor force. Income, Assets, and Debts Amount of income for each household member age 16+, gifts or loans between R and friends or relatives, householders or spouses only: value of home, real estate and vehicles, amount of savings and investments, debt amounts and payments, monthly rent. Abt Associates Inc. Chapter 7: A National Survey Of Families And Households 79 Exhibit 7.7 NSFH – Wave 2: Content of Older Focal Child Aged 18-23 Questionnaire Living Arrangements and History Current living situation, marital/cohabitation status, number of others in household, for each: name, sex, age, marital status, relationship to R, householder, (only if stepchild of R1 for whom may not have from R1 history); parent history: periods of six months or more not living with biological mother/father, ever lived on own for four months or longer, if lives with parent, how living with parent works out. Marital History Number of marriages; dates, first spouse and current spouse (if not first), second, third and fourth: cohabitation with spouse before marrying, cohabitation. Education High school diploma or currently attending high school, (If not HS grad) plan (expect) graduation or GED in next 12 months, expected educational attainment, grades, postsecondary education, grades, degrees and certificates/year each. Dating In past 30 days, times dated, number of different dates in past 12 months, steady boy/girlfriend, marriage plans. Fertility Fertility: number of children, for each: birth date and with whom living; currently pregnant, want child sometime, desired family size, feelings if nev er had a (another) child, how happy when first realized own pregnancy. Relationships with Mother Frequency contact, disagreements, arguing and fighting, global relationship, feelings about aspects of relationship, happiness with child's schooling, relati onships and occupation, likelihood of talking with parent, how influence. Age, Religion, and Economic Variables Birth date, religion, funding sources and expenses for school, active duty, employment history, current employment, occupational aspirations, spouse/partner’s employment, self/spouse income past 12 months, amount of savings, credit card or installment debt, other loans. Relationships with Father Frequency contact, disagreements, arguing and fighting, global relationship, feelings about aspects of relationship, happiness with child's schooling, relationships and occupation, likelihood of talking with parent, how influence, frequency do things with both parents. Partner Relationships If married: global marital happiness, stability; if cohabiting: likelihood will marry partner, global relationship with partner, stability. Social Support Leisure time with friends, how many persons could ask for help or advice (relatives other than parents, friends). Attitudes and Psychological Well -Being Gender roles, parenting, marriage, divorce, family obligations, self -esteem, global well-being; satisfaction with: school achievement, career prospects, financial situation, leisure time, friendships, health, love life, physical appearance, global health, height, weight, partial CESD depression scale, other depression items, efficacy. Sex and Drugs Sexual activity, cigarette, alcohol and marijuana use, ever live with problem drinker or drug abuser. Relations with Step-Parents Ever lived with stepparent, c ontact, disagreements, arguing and fighting, global relationship, Feelings about aspects of relationship, happiness with child's schooling, relationships and occupation, likelihood of talking with parent, how influence. Siblings and Grandparents, Interview Date Number siblings, relationship with siblings, contact and relationship with grandparents on mother’s side of family, contact and relationship with grandparents on father’s side of family, interview date. 80 Chapter 7: A National Survey Of Families And Households Abt Associates Inc. Exhibit 7.8 NSFH – Wave 3: Content of Interview with Main Respondent and Spouse Household Composition Detailed household composition. Children living elsewhere: names, distance, how often R sees them; martial status, educational status, number of children or who currently living with. Spouse/ partner’s children living elsewhere: martial status, educational status, number of children or who currently living with. Household Tasks Time spent by R, spouse/partner, others in household over 5 years of age. Goings and Comings of Children All children on previous and current household roster: lived outside household, number of times, when, reason, with whom did child live. Children on current but not previous household roster: date joined household, reason, lived outside household since joining. Children on previous household roster and current away roster: date living outside household, with whom did child live. Children on away roster at both previous and current interview: ever lived with R since that interview, when and why. Spouse/partner’s children on both previous and current roster: when began living apart, with whom did child live, occurrence of living in household. Spouse/partner’s children on away roster at both times: ever lived with R since then, number of times. Health Long-term physical and mental conditions of household members, activities of daily living (if over 55). Caregiving and Receiving Personal care given/received to/from member of household and to/from person outside household in last 12 months; most serious condition, person who gave/received most help. Relationship with Parents For mother, father, spouse/partner’s mother and father, step-parents: age or date died, health, relationship with, marital status, residence location; frequency of communication by phone, letter, or email; periodicity of visitation. For mother, father: live with sibling, live in nursing home or long-term care facility. Help and Support Given to and Received From Parents Help given/received (includes parents -in-law), which parents, how m any hours. Parent Health and Well-Being Hospitalization, help getting around, memory problems, too little income; which parents. Co-residence with Parent(s) R in parent’s household at previous interview: change in co-residence, date. R not in parent’s household at previous interview but is now: date co-residence began. R not living in parent’s household now: co-residence since previous interview. Help Given and Received Children 19 and older, includes R’s and spouse/partner’s children; others. Respondent’s Health and Well-being Overall happiness, satisfaction with various aspects of life, worry about income, depression and self-efficacy, alcohol and cigarette use, height and weight, positive aspects of psychological well-being. Siblings Distance, frequency of seeing; frequency of communication by phone, letter, or email. Grandparenting Age of grandchild(ren), frequency of overnight stay; frequency of communication by phone, letter, or email. Primary responsibility for raising grandchildren. Marriage and Cohabitation Changes in marital status since previous interview, cohabitation/marriage history; characteristics of partners/new spouses; for up to 3 marriages: likelihood of getting (re)married, dating and sex. Abt Associates Inc. Chapter 7: A National Survey Of Families And Households 81 Exhibit 7.8 NSFH – Wave 3: Content of Interview with Main Respondent and Spouse Recent Marital Disruptions Questions about end of marriage, dating, life now compared to before separation, contact with in-laws and former spouse, current relationship with former spouse, living arrangements. Marriage and Relationships (Unmarried, not Cohabiting) Life now compared to if were married; if steady girlfriend/boyfriend: satisfaction with relationship; views about marriage and relationships. Marriage and Relationships (Currently Cohabiting) Questions about current relationship, likelihood of marriage, satisfaction with relationship, fairness in relationship, time spent together, disagreements and arguments, life now compared to if separated, trouble in relationship, chances of relationship ending, views about marriage and relationships. Fertility History (skipped for females 55 and older and non-married men 55 and older) Biological or adopted children since previous interview: number, date of birth, planning status, if ever lived apart; currently pregnant: due date, planning status; future birth plans, sterilization operation, intention to have more children. Problem Inventory for All Children Age 5-17 Difficulty or ease in raising, long-lasting physical condition, mental or emotional problem; school: drop out, repeated a grade, suspended, expelled, skipped; in trouble with police, gotten (someone) pregnant while not married, seen a therapist for emotional problems. Problem Inventory for Children Age 18-23 Difficulty or ease in raising, long-lasting physical condition, mental or emotional problem; dropped out of school before completing high school, in trouble with police, gotten (someone) pregnant while not married, seen a therapist for emotional problems. Relationships with Young Adult Focal Children Time spent together, open disagreements, avoid talking about certain topics, global relationship, how disagreements handled, relationship quality, happiness with child’s schooling, relationships, occupation, likelihood of talking with child, how tries to influence child; if child lives in household: how it works out, pay for room or board, amount; expenses paid by R. Residential History Lived elsewhere; change of address (up to 5 different cities); number of moves. Education and Religion Religious preference, how often attend services, how religious; education history: school attendance since previous interview, types of school attended. Military Service and Work History If under age 45: active duty since previous interview, when; work history since previous interview, pensions and retirement. Current Employment Worked last week, hours worked, pay (“main job” if more than 1 job), kind of work, main activities, duties; if older than 60: kind of work doing at age 50; worked weekends, work unexpectedly, second job if more than one job. Work Experience in Past Year Weeks worked last year, weeks unemployed; if unemployed: earnings, kind of work, main activities at last job; if older than 60: kind of work doing at age 50. 82 Chapter 7: A National Survey Of Families And Households Abt Associates Inc. Exhibit 7.8 NSFH – Wave 3: Content of Interview with Main Respondent and Spouse Spouse/Partner’s Current Employment Weeks worked in past year, weeks unemployed, hours worked last week, full time/part-time status, income (household members over age 16): from wages and salaries in past 12 months, from other sources in past 12 months; whether R or others in household receive income from 7 other sources; public assistance: years received. Economic Relationship with Relatives Living in R’s Household If R owns house: does another randomly selected family member work, pay household expense, does R pay any of person’s expenses; if R lives in household of relatives - does R pay any household expenses. Home Purchase If purchased home since previous interview: when, amount paid. Inheritances Inheritances since previous interview, source, amount. Gifts and Loans from/to Persons not Living with R Gift given or loan received to/from people not living in household in past 12 months, from or to whom. Assets and Debts Mortgage payments, value of savings and investments and debts (amount owed each month). Family Attitudes, Social Participation and Work Family related attitudes, how often spend time in various social activities. Abt Associates Inc. Chapter 7: A National Survey Of Families And Households 83 Exhibit 7.9 NSFH – Wave 3: Content of Focal Child Interview Household Composition Living situation: dorm, military barracks; marital status, cohabitation status; if not living in group quarters: name and relationship of members of household. Separations from Biological Parents Ever separated from biological mother and father before age 18, number of times, how long, why, where lived, age at separation. Leaving Parent’s Home If parents not mentioned on current roster: when left first left parent’s home, distance from parents’ home, why left; if currently lives with parent: ever lived away, when left, when returned; other spells of leaving and returning home. Marital History If ever married: number of times, dates of marriages and separations, characteristics of spouses. Cohabitation History Dates of cohabitations and separations, characteristics of partners. High School Education If received high school diploma or G.E.D.: dates received; if still in high school: level; educational aspiration and grades received in high school. Post-Secondary Education Types of schools, dates attended each type, grades received, degrees received and dates, financial aid received, assistance from family, tuition cost. Dating Number of dates, dating partners, if steady boy/girl friend, how often see, satisfaction with relationship with boy/girl friend, views about marriage and living together. Fertility Biological or adopted children since previous interview: number of children, when born, planning status; ever adopted: number and when; currently pregnant, future birth plans. 84 Chapter 7: A National Survey Of Families And Households Abt Associates Inc. Exhibit 7.9 NSFH – Wave 3: Content of Focal Child Interview Relationship with Mother How often see; frequency of communication by phone, letter, or email; disagreements, arguing and fighting, global relationship, feelings about aspects of relationship, happiness with schooling, relationships, occupation, likelihood of talking with parent, how tries to influence. Employment Experience Military service and dates, dates of work his tory (when not primarily a student), periods of unemployment; if currently employed: hours work, occupation, rate of pay; if unemployed: looking for job, occupation at most recent job, rate of pay; if married or cohabiting: spouse/partner’s current employment; feelings about work. Income Income from wages, salaries and other sources in past 12 months for R and spouse/partner, savings, investments, debts, health insurance. Household Tasks Time R and other household members spend on household tasks. Relationship with Father How often see; frequency of communication by phone, letter, or email; disagreements, arguing and fighting, global relationship, feelings about aspects of relationship, happiness with schooling, relationships, occupation; likelihood of talking with parent, how tries to influence. Happiness with Marital Relationship Overall relationship quality and satisfaction with spouse, disagreements, trouble in marriage, likelihood of separation. Happiness with Cohabiting Relationship Overall relationship quality and satisfaction with partner, disagreements, trouble in relationship, likelihood of separation. Social Integration/ Social Support Leisure time with friends, friends and family can talk to. Religion Religious preference and attendance. Family Attitudes Health and Well-Being Life satisfaction, overall health and health indicators, depression, self-mastery and positive aspects of psychological well-being. Sexual Activity Age first had intercourse, number partners, frequency in past 30 days. Tobacco Use Age first smoked, number cigarettes per day in last 30 days. Alcohol Use Age at first drink, days drank in 30 days, number of drinks. Marijuana Use Age first used, number of times used, when last used. Problem Drinker in Family of Origin If yes: who. Abt Associates Inc. Chapter 7: A National Survey Of Families And Households 85 Exhibit 7.9 NSFH – Wave 3: Content of Focal Child Interview Relationship with Step-parent Ever lived with step-parent, how often see; frequency of communication by phone, letter, or email; disagreements, arguing and fighting, global relationship, feelings about aspects of relationship, happiness with schooling, relationships, occupation; likelihood of talking with step-parent, how tries to influence. Siblings Number siblings, distance of residence; how often see; frequency of communication by phone, letter, or email. Grandparents How often see; frequency of communication by phone, letter, or email. 86 Chapter 7: A National Survey Of Families And Households Abt Associates Inc. Exhibit 7.10 NSFH: Measures of Union Quality and Attitudes Union quality items (for married respondents; analogous items in most cases for cohabiting respondents): • Taking things all together, how would you describe your marriage? (Scale of 1 to 7 from “very unhappy” to “very happy”) • Were you married by a priest/rabbi/minister, judge/justice of the peace, no marriage ceremony? • How do you feel about the fairness in your relationship in each of the following areas: household chores, working for pay, spending money, child care? (5 point scale from “very unfair to me” to “very unfair to him/her) • During the past month, about how often did you and hour husband/wife spend time alone with each other, talking, or sharing an activity? • About how often did you and your husband/wife have sex during the past month? • How often, if at all, in the past year have you had open disagreements about each of the following: household tasks, money, spending time together, sex, having a(nother) child, inlaws, the children? • When you have a serious disagreement with your husband/wife, how often do you just keep your opinions to yourself, discuss your disagreements calmly, argue heatedly or shout at each other, end up hitting or throwing things at each other? • During the past year, have you ever thought that your marriage might be in trouble? Do you fell that way now? During the past year, have you and your husband/wife discussed the idea of separating? • It is always difficult to predict what will happen in a marriage, but realistically, what do you think are the chances that you and your husband/wife will eventually separate or divorce? General attitude towards marriage (for unmarried, non-cohabiting respondents): • I am not mature enough yet for marriage. • I would rather be able to spend time with my friends than be married now. • I would like to get married someday. • My parents would like it better if I were married now. • It would be all right for me to have children without being married. − Even if I had no plans to marry the father/mother. − If I had definite plans to marry the father/mother. • It would be all right for me to live with someone without being married. − Even if we had no interest in considering marriage. − To find out wither we were compatible for marriage. − If we were planning to get married. • I am not ready to have a sexual relationship with anyone. • I am not mature enough yet to live with someone. • I would like to live with someone before getting married. General attitude towards marriage (for cohabiting respondents): • I am not mature enough yet for marriage. • I would rather be able to spend time with my friends than be married now. • My parents would like it better if I were married now. • It would be all right for me to have children without being married. − Even if I had no plans to marry the father/mother. − If I had definite plans to marry the father/mother. Abt Associates Inc. Chapter 7: A National Survey Of Families And Households 87 Exhibit 7.10 NSFH: Measures of Union Quality and Attitudes How life would be different if married now (much worse to much better; for cohabiting and non-cohabiting unmarried respondents): • Standard of living • Economic security • Overall happiness • Freedom to do what you want • Economic independence • Sex life • Friendships with others • Relations with parents • Emotional security Considerations about timing of marriage (not at all important to very important; for cohabiting and noncohabiting unmarried respondents): • Having enough money saved • Finishing all of the schooling you plan to get • Being established in your job • Your partner finishing all of the schooling he/she plans to get • Your partner being established in his/her job Considerations about whether to marry someone: How willing you would be to marry someone who … (not at all willing to very willing; for cohabiting and non-cohabiting unmarried respondents): • Was older than you by 5 or more years • Was younger than you by 5 or more years • Had been married before • Already had children • Was not likely to hold a steady job • Was of a different religion • Was of a different race • Would earn much less than you • Would earn much more than you • Was not “good looking” • Had more education than you • Had less education than you Reasons for cohabitation (not at all important to very important; cohabiting and unmarried noncohabiting respondents): • It requires less personal commitment than marriage • It is more sexually satisfying than dating • It makes it possible to share living expenses • It requires less sexual faithfulness than marriage • Couples can make sure they are compatible before getting married • It allows each partner to be more independent than does marriage Reasons against cohabitation (not at all important to very important; cohabiting and unmarried noncohabiting respondents): • It is emotionally risky • My friends disapprove • My parents disapprove • It is morally wrong • It is financially risky • It requires more personal commitment than dating • It requires more sexual faithfulness than dating 88 Chapter 7: A National Survey Of Families And Households Abt Associates Inc. Exhibit 7.10 NSFH: Measures of Union Quality and Attitudes How your life would be different if you separated (much worse to much better; cohabiting and married respondents): • Your standard of living • Your social life • Your career opportunities • Your overall happiness • Your sex life • [If you have children] being a parent How your partner’s life would be different if you separated (much worse to much better; cohabiting and married respondents): • Your standard of living • Your social life • Your career opportunities • Your overall happiness • Your sex life • [If you have children] being a parent Abt Associates Inc. Chapter 7: A National Survey Of Families And Households 89 Chapter Eight: The Panel Study Of Income Dynamics The Panel Study of Income Dynamics (PSID) began in 1968 by selecting a nationally representative sample of 5,000 households. Since then, the study has followed the members of these households and their successors with annual interviews through 1996, and biennial interviews starting in 1997. The sample was supplemented in 1997 by the addition of 441 immigrant families (who were not in the 1968 sample frame). While the focus of the study is the dynamic aspects of economic and demographic behavior, it includes a wide range of sociological and psychological measures as well. Originally funded by the Office of Economic Opportunity, US Dept. of Commerce, the PSID is currently funded by the National Science Foundation. Additional funding has been provided by the National Institute on Aging, the National Institute of Child Health and Human Development, the Office of the Assistant Secretary for Planning and Evaluation of the United States Department of Health and Human Services, the Economic Research Service of the United States Department of Agriculture, the United States Department of Housing and Urban Development, and the United States Department of Labor. Survey Design The initial PSID sample consisted of two parts: an equal probability sample of 3,000 households from the 48 contiguous states, and a supplementary sample of 2,000 low-income households with heads under age 60, from standard metropolitan statistical areas (SMSAs) in the North and from nonSMSAs in the South, who had participated in the Survey of Economic Opportunity in the mid-1960s. In 1997 the core sample (descendent households of the orig inal 5,000) was reduced, and a refresher sample of post-1968 immigrant households with their adults children was added. Sample size: As of 2001, the PSID included over 7,000 families. It is expected to grow to 7,400 families by 2005. Methods of data collection: The survey was administered in person from 1968 to 1972, and thereafter primarily by telephone. CATI was introduced in 1993. By 1999, virtually all (97.5 percent) of interviews were CATI-based, and the remainder were CAPI-based. One member of the household, the “head of household,” is interviewed—the same individual from year to year as long as he or she remains the head. In a married couple family, this is the husband unless he is severely disabled. Response rates: The response rates in the first two years of the study were 76.0 percent and 88.5 percent. Since 1969, the response rate has ranged between 96.9 and 98.5 percent. As of 1988, the response rate for individuals who lived in 1968 households was 56.1 percent. Survey Content The content of the survey has remained largely unchanged over the decades in order to ensure comparability. The core topics addressed regularly in the PSID include income sources and amounts, poverty status, public assistance, other financial measures (taxes, inter-household transfers), family structure and transitions (marital events, birth and adoptions, children forming households), labor Abt Associates Inc. Chapter 8: The Panel Study Of Income Dynamics 91 force participation, housework time, housing, geographic mobility, socio-economic background (education, ethnicity, religion, military service, parents’ education, occupation, poverty status), and health. Supplemental modules have included: • • • • • • • • • • • • • Housing and neighborhood characteristics (1968-1972, 1977-1987) Achievement motivation (1972) Child care (1977), child support and child development (1997) Job training and job acquisition (1978) Retirement plans (1981-1983) Health: health status, health expenditures, health care of the elderly and parent's health (1986, 1990, 1991, 1993-1995) Kinship: financial situation of parents, time and money help to and from parents (1980, 1988) Wealth: assets, savings, pension plans, fringe benefits (1984, 1989, 1994, 1999) Education: grade failure, private/public school, extracurricular activities, school detention, special education, Head Start Programs, criminal offense (1995) Military combat experience (1994) Risk tolerance (1996) Immigration history (1997) Time use (1997) Retrospective marriage and fertility histories were obtained in 1985. In 1997 a Child Development Supplement was added, focusing on children aged 0 to 12 in PSID households. It collects information on home, school, and day care environments, and cognitive, emotional, and physical functioning. Addition of census geocodes has allowed incorporation of neighborhood characteristics from the decennial censuses. Suitability for Research Outcome measures: The PSID collects data on union status at each interview and collects information explicitly about marriages, separations, and divorces between interviews (month and year of occurrence). Since 1983, relationship codes distinguish between the head’s wife, “wife” or cohabitor, and girlfriend—the distinction between the latter two being that a “wife” or cohabitor has 92 Chapter 8: The Panel Study Of Income Dynamics Abt Associates Inc. been living with the head for at least a year.19 Additionally, retrospective marital and fertility histories were collected in 1985 to obtain information about changes in marital or childbearing status prior to the initial 1968 interview. Approximate cohabitation histories can be constructed from the Relationship File, which shows the relationship between each pair of individuals for each year from 1968 through 1985 at the time of the survey (but not the exact dates at which cohabitation transitions occurred). Virtually no information is collected on union quality. Measures of demographic influences: The PSID includes explicit marital and fertility histories, and offers the ability to construct an approximate cohabitation history. These histories permit the study of the influences of both early and non-marital childbearing on union formation and dissolution; PSID offers a particularly extended period of time to track any longitudinal influence of these earlier behaviors. The lack of satisfaction and interaction measures poses a barrier to using these data for studying transitions to parenthood, despite the availability of fertility histories. Basic information is available for studying effects of cohabitation on marriage, but not in the context of relationship quality, time spent together, values or expectations. Both biological parents of children are identified. Measures of economic influences: For both the household head and spouse or partner, labor force information includes work status, union membership, occupation, duties, type of business, salary/wage, months and hours worked, duration of job, and promotions. Educational attainment and job histories are also available. These features suggest the potential of these data for exploring effects of both men’s and women’s employment on union behaviors. Data also are available on household income sources and amounts, inter-household transfers, wealth and active savings, and expenses. The geographic identifiers, combined with the extensive individual-level economic information, make this an excellent data set for examining the role of external policy influences (such as AFDC, PRWORA etc.), as well as local labor and marriage markets. Measures of socio-cultural influences: Since the PSID does not collect any attitudinal data, the influence of individuals’ views on marriage, cohabitation, or gender roles cannot be examined directly. However, the measures of male and female employment and earnings available in the PSID, as well as fertility and household structure data and number of hours spent on housework, have the potential to describe men and women’s breadwinning and household labor behaviors and link these with union behaviors. This may offer indirect evidence of the influence of attitudes and expectations about roles on union outcomes. Measures of psychological influences: The PSID is an inappropriate data set for direct examination of psychological factors in union outcomes, as it does not collect data on relevant measures such as behaviors within relationships, feelings and emotions, or relationship satisfaction. 19 The head is typically male. The PSID website remarks: Originally, if the family contained a husband-wife pair, the husband was arbitrarily designated the Head to conform with Census Bureau definitions in effect at the time the study began. When a new head must be chosen due to changes in family composition, etc., it must be someone who is at least 16 years old and the person with the most financial responsibility for the family unit. But if that person is female and has a husband or a boyfriend of at least one year’s standing living with her, then the husband or boyfriend becomes the head. Abt Associates Inc. Chapter 8: The Panel Study Of Income Dynamics 93 Special strengths: The PSID is an extraordinary source of longitudinal data, having followed the same families and their spin-offs for 35 years. A wide array of information is available through the topical modules. The extensive measures of income and employment, educational attainment, and family background are useful for developing alternate approaches to the definition of disadvantage, and examining associated variations in union formation and dissolution. The recent inclusion of a substantial immigrant sample offers opportunities to study union behaviors for this diverse and often disadvantaged population. Limitations: For the past five years, data have been collected only biennially, and prior to that, annually. Hence the PSID does not have the fine level of detail found in the CPS and the SIPP, which share a similar topical emphasis on economic well-being. PSID’s strengths in measuring economic influences contrasts sharply with the total absence of attitudinal or psychological measures. Sample attrition, and associated issues of generalizability and representativeness, may limit the usefulness of some lines of inquiry, particularly among the later spin-off generations. Using the Data The PSID website is http://www.isr.umich.edu/src/psid/. Documentation: Technical documentation, including a user’s guide, is available online at: http://www.isr.umich.edu/src/psid/doc.html. Data Availability: With the exception of some of the special supplemental files, data can be accessed through the online PSID data center. Codebooks are also available for downloading. The website is: http://simba.isr.umich.edu. Reports and Articles: A list of publications, papers and reports based on PSID data can be found at http://www.isr.umich.edu/src/psid/bibliography/biblio.html. Some relevant studies using PSID data include: Ellwood, D. (2000). The impact of the earned income tax credit and social policy reforms on work, marriage, and living arrangements. National Tax Journal, 53(4), 1063-1106. Hoffman, S. D. & Duncan, G. J. (1995). The effect of incomes, wages, and AFDC benefits on marital disruption. Journal of Human Resources, 30(1), 19-41. Ono, H. (1998). Husbands’ and wives’ resources and marital dissolution. Journal of Marriage and the Family, 60(3), 674-689. Data Elements The contents of the core survey (2001) are described in Exhibit 8.1. Major supplemental topics over the years are listed in Exhibit 8.2, and special supplemental files in Exhibit 8.3. 94 Chapter 8: The Panel Study Of Income Dynamics Abt Associates Inc. Exhibit 8.1 PSID: Core Survey, 2001 TN: Thumbnail Sketch Non-family unit household members, potentially confusing family situations, ambiguous or conflicting information, anything else editors / coders should be aware of, institutional housing situation. Birth Dates and Age Calculations Verify birth date information, deceased family members: verify birth date, birth location, death date / location. Gender, marital status, name of spouse. Marital Status and Checks Marital status, name of spouse. Parent / Child Relationships Verification of family relationships for new members of family: natural / adopted child, child’s name / age, status of child’s parents in family unit, elsewhere, deceased, don’t know), identify parents for each child in family unit. Split-off Sample Creation Family unit member no long in same household: address and phone number, marital status, name of spouse, own household / institution. Child moved. Move-in Date and Checks – Reinterview and Recontact Date new family unit member moved in, date family unit member(s) moved out/died. Move Out Date and Checks – Reinterview Date family unit member(s) moved out/died. A. Housing Type of dwelling. If household head in retirement facility, what services received, cost of services. Method of heating home, size of home, heating subsidy, home owned/rented. If owned: value, taxes, insurance premiums, mortgage / second mortgage / other loans, down payment details, experience with mortgage lender, experience obtaining other loans, experience with rejected loan applications. If home is rented or rented by non-family unit member: amount paid, furnishings/heat included in rent, public housing status, government subsidized rent. Steps taken towards home ownership. Air conditioning, most recent move, chance of future move. B. Employment of Head of Household Work status, self-employment status, business ownership, information on employer, union membership / labor contract, occupation, work duties, type of business/industry, salary/wage earned, overtime pay / tips / commission, job searching activities, duration of time with current employer (start date), promotions, months / hours worked, work missed due to: sickness (of self or others), vacation taken, strikes, unemployed or laid off. Second, extra job(s). C. Head is not Working Now Currently looking for work, job searching activities, last job held: date (start/end), duration, position, duties, type of business, self-employment status, multiple jobs, reason for end, changes in work position/ promotion, starting salary/wage, hours/week, multiple jobs held at once (questions asked for each job held). Work missed due to: sickness (of self or others), vacation taken, strikes, unemployed or laid off. D. Employment of Wife / “Wife” Work status, self-employment status, second / multiple jobs, information on employer, union membership/labor contract, occupation, work duties, type of business/industry, salary/wage, overtime pay / tips / commission, job searching activities, duration of time with current employer (start date), change in position / promotions, months / hours worked, work missed due to: sickness (of self or others), vacation taken, strikes, unemployed or laid off. Second, extra job(s). E. Wife / “Wife” is not Working Now Currently looking for work, job searching activities, last job held: date (start/end), duration, position, duties, type of business, self-employment status, multiple jobs, reason for end, changes in work position/ promotion, starting salary/wage, hours/week, multiple jobs held at once (questions asked for each job held). Work missed due to: sickness (of self or others), vacation taken, strikes, unemployed or laid off. Abt Associates Inc. Chapter 8: The Panel Study Of Income Dynamics 95 Exhibit 8.1 PSID: Core Survey, 2001 F. Housework, Child Care and Food Costs Hours / week wife spends on housework, hours / week family head spends on housework, number of family sit down meals in week, child care costs, food stamps (value / months used), money spent on food (for home / delivered / eating out), food shortage. G. Income Farm or Ranch: total receipts from farming, total operating expenses, net income. Business ownership or financial interest in a business enterprise: number/type of businesses, owner, self or wife work for business, corporation status, profitability, share of income / loss, wages / bonuses/ overtime / tips / commission earned from business; income from professional practices or trade, farming or market gardening, roomers or boarders, rent, dividends, interest, trust funds or royalties, ACD/AFDC, SSI, welfare, Social Security, Veteran’s Administration (servicemen’s / survivor’s pension, service disability, GI bill), retirement pay, pensions, annuities, unemployment compensation, workers compensation, child support, alimony or separate maintenance, help from relatives, other people not living with you and not your family, income from anywhere else (amount, months received). Wife’s income, earnings from work (before taxes), unemployment compensation, worker’s compensation, income from dividends, interest, trust funds or other assets, SSI, ADC/AFDC, child support or other welfare, pensions or annuities, financial help from relatives or friends, other income. Household members who moved out or died. Household roster (relationship to head, name), income of other household members from work, interest, SSI, welfare, Social Security, other sources. Receipt of insurance settlement / inheritance, itemized tax deductions, money for support of non-household members, reasons for money shortage, debt/loan consolidation, creditors demanding payment, experience with wages attached, lien filed against property, property repossessed, bankruptcy filing. H. Health and Healthcare Costs Questions asked about self and wife: overall health, limitations on type of work possible due to health, health conditions (stroke, hypertension, diabetes, etc), difficulty in caring for self (bathing, dressing, eating, getting in/out of bed, walking, using toilet), light/vigorous physical activity, cigarettes, alcohol, height and weight, feelings of sadness/nervousness/restlessness, members of family covered by Medicaid or medical assistance (Medi-Cal), type of health insurance or coverage, out-of-pocket medical and dental expenses, total cost of family medical care. J. Marriage and Fertility History Number of marriages, date of marriage, spouse’s name, status of marriage (widowed/ divorced/ separated; date). Number of children, names, gender, birth dates, birth weight, birth state, location of child now, if dead, date of death, race, primary / secondary ethnic group. Legally adopted children, number, names, gender, birth date, relationship to R before adoption, birth state, location of child now, if dead, date of death, race, primary / secondary ethnic group. K. Background and Education of New Wife / “Wife” Father’s education / ability to read and write, mother’s education/ ability to read and write, number of siblings, number of siblings still living, ethnic origin / race, participation in military, education, high school graduation / GED attainment, grades of school finished, las t year in school, college, last year in college, highest year of college completed, highest degree received, name / location of college, degree through a vocational / training / apprenticeship program, religious preference / denomination, number of years worked for money since age 18. L. Background and Education of New Head Parent’s place of origin, parent’s occupations, first full-time job, variation in jobs, setting / location of childhood home (rural / suburban / urban), experience living in other states / countries, experience moving to new community, SES of parents during childhood, father’s level of education / ability to read and write, mother’s level of education / ability to read and write, number of siblings, number of siblings living, death of parent during childhood, ethnic origin / race, participation in military, education, high school graduation / GED attainment, grades of school finished, last year in school, college, last year in college, highest year of college completed, highest degree received, name / location of college, degree through a vocational / training / apprenticeship program, religious preference / denomination, number of years worked for money since age 18. 96 Chapter 8: The Panel Study Of Income Dynamics Abt Associates Inc. Exhibit 8.1 PSID: Core Survey, 2001 M. Risk Aversion Series of questions asked to assess risk aversion. Questions deal with willingness to take risks for increasing family income. P. Pensions Pension or retirement plan, years included, vested, optional/required contributions, employer contributions, investment of retirement funds, age for receiving benefits, early retirement option, designee to receive funds in case of death, pension dependent on SS, typical retirement age for job, retirement plans, thrift/profitsharing/Keogh plan, employer contributions, plans through former employer, type of plan (formula / accumulation), age / date when benefits began/will begin, benefits adjusted for cost of living. Cash settlement, IRA rollover, expected future benefits. R. Welfare Reform Public assistance or welfare payments from state/local welfare office: who in family unit, state of residence at time, type of assistance, amount, months received. SSI / Child Support / Help from relatives / help from someone else: who in family unit, amount received, months received. Total income for family unit, comparison of income to previous year, employment during year, income from job, number of weeks worked, months employed, hours / week, months unemployed. Since January 1999: stoppage of benefits / food stamps being received, when/why, reapply for public assistance, family members required to work/attend school/ other requirements for pubic benefits, complete application for public assistance, which programs, application approved/denied/ pending, received any of the following types of assistance (from government, church, family or community group)– transportation, housing, child care, health care, job training, food, clothing, other expenses. T. Philanthropy Donations of money/assets/property (>$25): to religious/charitable organization, to combination purpose organizations (United Way, United Jewish Appeal, etc.), to organizations that provided food/ shelter/basic necessities, to health care or medical research organizations, to education purposes, to other charitable organizations, total dollar amount of donations. Volunteering: hours spent on: helping provide food / shelter / basic necessities. V. Vehicles Inventory Own / lease car for personal use, number of vehicles. All vehicles: size, primary driver, purchased / leased / gift, new/used, year and make, date acquired, used (primarily) for business purposes (other than to/from work), used primarily for disabled transportation. If purchased: trade-in or sell another vehicle at time, down payment amount, amount borrowed / financed, payments (amount and periodicity), number of payments to be made, number of payments already made. If leased: initial lease payment, payments (amount and periodicity), number of payments to be made, number of payments already made. W. Wealth and Active Savings Value (minus debts / money still owed) of: real estate other than main home, vehicles (including boats / trailers), farm/business value of stocks / mutual funds / investment trusts, annuities / IRAs, checking/savings accounts, money markets, CDs, bonds, treasury bills, bond funds, cash value in a life insurance policy, valuable collection, rights in a trust or estate. Debts: credit card, student loans, medical/legal bills, loans from relatives. Changes in wealth. Since January: put money in annuity/IRA, pension, sell main home / secondary property, acquire property, make substantial (>$10,000) improvements/additions to home, money invested in farm/business, sell part/all of farm/business, buy / sell stocks, mutual funds, investment trusts, person leaving family who took $5000 or more in assets or debts away, person joined the family who brought $5000 or more in assets or debts. In last 2 years: gifts/inheritances of $10,000+. Section X. Expenses Car insurance, car payments, transportation related expenses (repair / maintenance, gas, parking, car pool, bus/train fares, cabs, other transportation costs), school-related expenses (books, supplies, uniform, computer, tuition, tutoring, room and board for family member away at school). Abt Associates Inc. Chapter 8: The Panel Study Of Income Dynamics 97 Exhibit 8.2 PSID: Major Supplemental Topics Housing and neighborhood characteristics (1968-1972, 1977-1987) Achievement motivation (1972) Estimating risk tolerance (1996) Child care (1977) Child support and child development (1997) Job training and job acquisition (1978) Retirement plans (1981-1983) Health—health status, health expenditures, health care of the elderly and parent's health (1986, 1990, 1991, 1993-1995) Kinship—-financial situation of parents, time and money help to and from parents (1980, 1988) Wealth—assets , savings, pension plans, fringe benefits (1984, 1989, 1994, 1999) Education—grade failure, private/public school, extracurricular activities, school detention, special education, Head Start Programs, criminal offense (1995) Military combat experience (1994) Risk tolerance (1996) Immigration history (1997) Time use (1997) Exhibit 8.3 PSID: Special Supplemental Files Active Saving File, 1984-1989 and 1989-1994 Childbirth and Adoption History File, 1985-1997 Death File, 1968-1997 Ego-Alter File, 1985 (retrospective data on substitute parenting, use of public assistance) Estimating Risk Tolerance for the 1996 PSID Family “Income Plus” Files, 1994-1997 Geocode Match File Health Care Burden File, 1993 Hours of Work and Wage Files, 1994-1997 Marriage History File, 1985-1997 Medicare File OFUM (“Other Family Unit Member”) Income Detail File, 1993 Parent Health Supplement, 1991 Parent Identifier File, 1968-1997 Relationship File, 1968- 1985 Retrospective Occupation-Industry Files, 1968-1980 Self-Administered Health Supplement, 1990 Telephone Health Supplement, 1990 Time and Money Transfers File, 1988 Wealth File, 1984, 1989, 1994, 1999 Work-History File, 1984-1985; 1984-1986; 1984-1987 98 Chapter 8: The Panel Study Of Income Dynamics Abt Associates Inc. Chapter Nine: The Survey Of Income And Program Participation The Survey of Income and Program Participation is a nationally representative survey of the noninstitutionalized US population that has been in continuous operation for nearly 20 years. It collects a wide variety of economic and demographic information on panels of respondents over a period of several years, contacting sample members every four months. It is conducted by the Bureau of the Census. Survey Design Each SIPP panel of households is derived through a two-stage process: selection of primary sampling units (PSUs), which are counties and independent cities; and selection of address units within PSUs. Sample members are the residents at those addresses. Information is also collected on individuals who join their households through birth or moving in. Original sample members who are aged 15 or over are followed if they move away. The sampled households are randomly divided into four rotation groups, which are interviewed in successive months. The number of interviews per panel and the spacing of the panels have varied over the life of the SIPP. From 1984 through 1993, new panels were introduced annually, and sampled households were typically interviewed 8 or 9 times, i.e. covering about 2 ½ years. Some panels were truncated, however, due to insufficient funds: the 1988 Panel had six waves, and the 1989 Panel ran for only three waves before being folded into the 1990 Panel. (The 1992 panel, on the other hand, had 10 waves of interviews.) Consecutive panels overlapped, so that information for a particular calendar year could be obtained by combining appropriate waves of several panels. A major redesign in the mid 1990s was intended to lead to the replacement of short, overlapping panels with longer, abutting ones, with larger sample sizes. The 1994 and 1995 panels were dropped, and the 1996 panel was designed to run for four years, or 12 waves. The following panel began in 2001 rather than 2000, however, and will comprise only nine waves. Sample size: The 1992 and 1993 panels had about 20,000 households each. Although the original intention was that all panels be this size, lack of funds often led to sample size cuts. The 1991 panel, for example, contained only 14,000 households. The recent redesign calls for larger samples. The 1996 panel had an initial sample size of 40,188 households and the 2001 panel comprises some 37,000 households. The redesign also calls for oversampling of households from high poverty areas. Methods of data collection: Through 1991, interviews were generally conducted in person. Starting in 1992, interviews have been conducted by telephone where possible, as a cost saving measure. The Wave 1, 2, and 6 interviews are still done in person, however. Pencil-and-paper instruments were used through the 1993 panels. The SIPP redesign include the introduction of computer-assisted interviewing, which has been in use for the 1996 and 2001 Panels. Response rates: For the 1996 panel, non-response in Wave 1 was 8.5 percent. The sample loss by the end of the final wave was 35.5 percent. Abt Associates Inc. Chapter 9: The Survey Of Income And Program Participation 99 Survey Content The SIPP interview comprises three components: the control card, the core questionnaire, and topical modules. The control card contains information about the type of housing and the household roster with basic demographics (date of birth, race/ethnicity, sex, and education). The relationship of each household member to the reference person is shown, and additional variables identify sample members’ spouses and parents when they are in the same household. The core questionnaire includes sections covering labor force participation, earnings, sources and amounts of unearned income, assets, health insurance, participation in various assistance programs, and education activities. This information is collected for all members aged 15 and older. Topical modules vary by wave. The modules administered to the 1996 panel were: • • • • • Wave 1: assistance recipiency history, employment history Wave 2: marital history, fertility history, work disability, education and training history, migration history, household relationships Wave 3: assets, liabilities, and eligibility; medical expenses/utilization of health care, work-related expenses, child support paid Wave 4: annual income and retirement accounts, taxes, work schedule, child care, disability Wave 5: school enrollment and financing, child support, support for non-household members, functional limitations, employer-provided health benefits, OMB welfare reform question Wave 6: children’s well-being; assets, liabilities, and eligib ility 20 ; medical expenses/utilization of health care, work-related expenses, child support paid Wave 7: annual income and retirement accounts, taxes, retirement expectations and pension plan coverage, home health care Wave 8: adult well-being, welfare reform Wave 9: assets and liabilities; medical expenses/utilization of health care, work-related expenses, child support paid Wave 10: annual income and retirement accounts, taxes, work schedule, child care Wave 11: child support, support for non-household members, functional limitations and disability Wave 12: assets, liabilities, and eligibility; medical expenses/utilization of health care, work-related expenses, child support paid, children’s well-being • • • • • • • 20 This module is so named because it includes information used for determining eligibility for some meanstested programs. Chapter 9: The Survey Of Income And Program Participation Abt Associates Inc. 100 Suitability for Research Outcome Measures: Information on union status is collected at each contact. Marital (but not cohabitational history) is collected as well. Changes in household composition from month to month can be used to construct a cohabitational history, limited to the length of the panel. No data are available on union quality. Measures of demographic influences: The SIPP includes complete marital and fertility histories, and reports the relationships among all members of the household. The marital and fertility histories permit the study of the influences of both early and non-marital childbearing on union formation and dissolution. However, most of the relevant covariates are collected only for the 3-4 year period of data collection, and the relatively short period of longitudinal data collection limits the usefulness of this survey for studying any long-term influences. Some aspects of the influence of the transition to parenthood can be examined, through the use of fertility histories or for individuals experiencing births during the survey period. The lack of satisfaction and interaction measures limits the usefulness of the SIPP for studying transitions to parenthood. The influence of cohabitation on marriage can be examined only for union transitions occurring during the period of data collection. Measures of economic influences: Detailed information is available on earned and unearned income, program participation, assets and liabilities, and taxes. Labor force measures include hours worked per week, wage/salary, union membership, employer size and location, occupation, industry, work activities and duties. Topical modules provide information on education and training, migration, and recipiency histories, work schedules, work-related expenses, and work disabilities. The recipiency history and detailed information about program participation and eligibility make the SIPP particularly relevant to studying disadvantaged populations. While the job and recipiency histories could be linked to the marital histories, the SIPP is more appropriate for examining causal relationships between more current economic influences and prospective measures of union transitions during the survey period. Measures of socio-cultural influences: The SIPP is a poor data set for direct examination of socio- cultural influences on union behaviors, as it does not collect data on attitudes, expectations, or aspirations. Measures of psychological influences: The SIPP is an inappropriate data set for direct examination of psychological factors in union outcomes, as it does not collect data on relevant measures such as behaviors within relationships, feelings and emotions, or relationship satisfaction. Special strengths: The SIPP collects a vast amount of economic and demographic information on sampled households over a period of up to four years. The longitudinal nature of the survey supports estimates of fixed effects models. The data are collected on a sub-annual basis. Information is collected on all members, and they are followed to new locations if they move away either individually or as a group. The topical modules obtain a complete map of the relationships among all household members, as well as life histories of marriage, fertility, employment, welfare recipiency, and migration. The extremely large sample size permits focus on detailed subgroups, which is particularly useful for extending research around issues of disadvantage. The emphasis on program participation is of particular note for the 1996 panel, which occurs during a period of substantial exogenous changes in public programs potentially relevant to union behaviors. Abt Associates Inc. Chapter 9: The Survey Of Income And Program Participation 101 Limitations: The SIPP does not collect any measures of attitudes, expectations, or aspirations, or descriptors of the quality of relationships. Cohabitation information is limited to measures during the survey period. While it has strong advantages over other cross-sectional surveys, it is limited compared to other longitudinal surveys in its ability to examine causal influences over an extended period. Using the Data The SIPP website is: http://www.bls.census.gov/sipp/. Documentation: Key reports for users are available on the SIPP website. These include: Survey Of Income and Program Participation Users’ Guide (Supplement to the Technical Documentation), Third Edition, Washington, D.C., 2001, U.S. Department Of Commerce Economics and Statistics Administration, U.S. Census Bureau, as well technical documentation specific to each data file. Data availability: All public use microdata files are available on magnetic media or CD-ROM, along with a full set of documentation, directly from the Census Bureau. Available files include core data by wave, topical module by wave, and full panel longitudinal files. Data extracts can also be downloaded directly from the SIPP website, via Surveys-on-Call (longitudinal files for the 1988 through 1993 panels, and individual wave and topical module files for the 1990 through 1993 panels) or FERRET (1996 panel and longitudinal files for the 1992 and 1993 panels). Reports and articles: The SIPP bibliography (on their web site) provides a list of published and unpublished documents related to SIPP. It includes studies using SIPP data and citations to methodological research about SIPP. Some examples are: Dickert-Conlin, S. (1999). Taxes and transfers: Their effects on the decision to end a marriage. Journal of Public Economics, 73, 217-240. Fitzgerald, J. M. (1991). Welfare durations and the marriage market: Evidence from SIPP. Journal of Human Resources, 26, 545-561. Data Elements The content of the core questionnaire and the topical modules is described in Exhibits 9.1 and 9.2, respectively. 102 Chapter 9: The Survey Of Income And Program Participation Abt Associates Inc. Exhibit 9.1 SIPP: Content of Core Survey for all Waves Section A: Instrument Front Questionnaire Race, gender, household size, status of living quarters (owned or rented); subsequent to Wave1: if any household members moved to a new location, date, reason and new address of individuals interviewed. Section B: Coverage Items Questionnaire Home and mailing addresses (if different), age of home, extra units in building or on property (vacant or occupied), determination of group living quarters, type of housing, home ownership status, public housing, receive a rent subsidy, Section 8 information. Section C: Household Demographics Questionnaire Household roster (name, usual residence, gender, relationship to reference person and to others in household, including step and adoptive relationships), birth dates, age, marital status, past and present; divorce/widowhood, participation in armed services, highest level of education, HS graduation/GED, are parents of household members also in household; adopted/step children, race, origin and descent, Social Security Number (SSN) or Railroad Retirement Number (RRN). Subsequent to Wave 1: Wave 1 household members no longer in household (date, reason, new address), household members new since Wave 1 (date, reason), change in person / people who own/rent home, changes in marital status. Section D: Labor Force – Part 1 Questionnaire Recent work history, reasons for not working, ability to work, receipt of unemployment pay (Workers’ Compensation, Unemployment Insurance, strike pay, union benefits), status of job search, date of lay off / expected return to work, number of employers in / before reference period, type of work, duration of work, business ownership, details of business, dates of business existence, reasons for end of business, hours worked per week, hourly rate, pay period, union membership / union contract, employer size and location(s), absence from work (without pay), support from social services for job search or training. Subsequent to Wave 1: Living situation during Wave 1 interview (if not interviewed in Wave 1), change in work / business duties since Wave 1. Section E: Labor Force – Part 2 Questionnaire Gross income received from job(s) (amount before deductions, pay rate / hours worked); repeated for all sources of earned income, other earnings, income from business owned by respondent. Section F: General Income – Part 1 Questionnaire Income sources since reference month, severance, pension or retirement from job left, Department of Veteran Affairs, Social Security for self or children, SSI for self or children, income because of health condition, specific sources, retirement income, life insurance or other annuity income, income as result of being survivor, foster care payments, child support / alimony payments, food stamps, WIC, state or county cash or other assistance for self and/or children in household, community service or job training activities. Assets Section, Part 1 Questionnaire Detailed list of financial investments (IRA, Keogh, 401K / thrift plan, interest earning checking account, savings account, money market, CD, mutual funds, stocks, bonds, US Government securities, mortgages, rental property, royalties, other financial investments, receipt of lump sum or regular distribution payments from financial investments). Section G: General Income – Part 2A Questionnaire Reasons for sources of income, disability payments (date / age at time of first payment), receipt of income sources jointly with spouse, type of Veterans payment, state / local welfare payments, WIC, Food Stamps (reasons for applying, date of application, reasons for stoppage of public funds, amount received, household members covered by funds / non-cash assistance received). Section H: General Income – Part 2B Questionnaire Income received, reinvestment or roll-over of money into IRA or other retirement plan (amount, who payments cover), assistance received for child(ren) (date begun, amount, who is covered). Assets Section, Part 2 Questionnaire Interest from assets (sole / joint ownership; total amount of interest earned), asset value, dividend checks (received solely / jointly; amount per year), rental property income (received solely / jointly; amount, net loss or gain), mortgage (sole / joint ownership, interest paid), royalties (amount of individual’s share), other investments (income received). Abt Associates Inc. Chapter 9: The Survey Of Income And Program Participation 103 Exhibit 9.1 SIPP: Content of Core Survey for all Waves Section J. Health Insurance Questionnaire Medicare (months covered, claim number), Medicaid (months covered, children covered, date begun), other public medical care payment program (months covered, children covered), SCHIP, health insurance plan (whose name on plan, plan obtained through, amount of premium covered, others on plan, coverage of other household / non-household members), reasons for not having health insurance coverage, last time without coverage. Section K. Programs Questionnaire Public / subsidized housing (date applied, date of move-in, on waiting list), amount paid in rent, utilities (water, electricity, gas / oil), energy assistance (form, amount), school lunch (offered, free/reduced qualified), school breakfast (offered, free/reduced qualified). Section L. Education Questionnaire School enrollment (full/part time, grade level), highest level / degree completed, educational assistance / financial aid (type). Section O. Instrument Back Questionnaire Telephone (in home, elsewhere to reach members of household), logistics for future interviewing. 104 Chapter 9: The Survey Of Income And Program Participation Abt Associates Inc. Exhibit 9.2 SIPP: Content of Topical Modules, by Wave Wave 1 Recipiency History. Obtains a profile of a respondent's pattern of participation in certain government programs prior to the beginning of the SIPP panel. Specific questions address the first time a respondent participated in a particular program, the length of participation, and the number of times the respondent has been in the program. [Panels and waves: 86-2, 87-2, 88-2, 89-2, 90-2, 91-2, 92- 1, 93-1, 96-1, 01-1] Employment History. Identifies patterns of employment, length of employment at certain jobs, and reasons for any periods of unemployment subsequent to the respondent's first job. Beginning with the 1996 Panel, specific questions that address type of work done, job duties, and the industry in which the respondent works were moved into the core content; previously, such questions had been part of this module. [86-2, 87-2, 88-2, 89-2, 90-2, 91-2, 92-1, 93-1, 96-1, 01-1] Wave 2 Work Disability History. Asks a series of questions about chronic health conditions that may affect the amount or type of work a respondent can do. Included are any such physical, mental, or other health conditions that interfere with the respondent's ability to work for at least 3 months. Questions are asked about when the limiting condition first became an issue, whether the person was working at the time, whether the condition resulted from an accident or injury, and if so, where the accident or injury occurred. Shorter-term conditions (including pregnancy) are not included as limiting conditions. [86-2, 87-2, 88-2, 89-2, 90-2, 91-2, 92-2, 93-2, 96-2] Education and Training History. Collects information about respondent's highest level of school completed or degree received, courses or programs studied, and dates of receipt of high school and postsecondary degrees or diplomas. The module determines if the respondent attended a public or a private high school. Job-related-training questions address training designed to help find or develop skills for a new job as well as to improve skills at the current or most recent job. People 15 years of age and older are asked whether they have received job training; if they have, they are asked about the duration of the training, how it was used, how it was paid for, and if it was federally sponsored.4 (Variations are also asked as Education and Work History [84-3] and Education and Training [84-6].) [86-2, 87-2, 88-2, 89-2, 90-2, 91-2, 92- 2, 93-2, 96-2, 01-2] Marital History. Asks questions of all respondents aged 15 and older who have ever been married. The date of the present marriage is determined; for those married more than once, SIPP records the dates of their first two marriages and their last marriage, if married more than twice. If appropriate, respondents are asked when their previous marriages ended and whether they were widowed or divorced at the end of their marriages. [848, 85-4, 86-2, 87-2, 88-2, 89-2, 90- 2, 91-2, 92-2, 93-2, 96-2, 01-2] Fertility History. Asked only of females aged 15 years and older and males aged 18 and older. Men are asked about the number of children they have fathered, and women are asked about their birth histories. Interviewers ask women who have had children when their first and last children were born, along with questions about their employment status during pregnancy and prior to the birth of their first child, circumstances of any absence from work before and after the first birth, and the maternity leave policies of their employers. Pos tbirth employment is also covered. [84-8, 85-4, 86-2, 87-2, 88-2, 89-2, 90-2, 91-2, 92-2, 93-2, 96-2, 01-2] Migration History. Asks respondents aged 15 and older where they were born, where they have lived, and how long they have lived in those places. Respondents born in a foreign country are asked about their citizenship status and when they came to the United States to stay. [84-8, 85-4, 86-2, 87-2, 88-2, 89-2, 90-2, 91-2, 92-2, 93-2, 96-2, 01-2] Household Relationships. Collects information about relationships among household members. The SIPP core questions gather extensive information about household composition for each month of the panel. This information allows for the identification of families and subfamilies and details each household member's relationship to the household reference person. As extensive as this information is, it does not cover the interrelationships of all household members. For example, the SIPP core provides no information about the relationships between members of two different unrelated (to the household reference person) subfamilies residing in the same household. This topical module fills that gap by collecting complete information about how each member of the household is related to every other member of the household. Relationships are specified in detail; for example, a brother is a full brother, half brother, stepbrother, or adoptive brother. In-law relationships are also identified. [84-8, 85-4, 86- 2, 87-2, 88-2, 89-2, 90-2, 91-2, 92-2, 93-2, 96-2, 01-2] Abt Associates Inc. Chapter 9: The Survey Of Income And Program Participation 105 Exhibit 9.2 SIPP: Content of Topical Modules, by Wave Wave 3 (repeated in Waves 6 and 9) Medical Expenses and Utilization of Health Care. Payment of housing, food and living expense. The following questions asked about respondent and about respondent’s child(ren): overall health, overnight hospitalization, prescription medication, visits to dentist, questions about adult teeth, visits to doctor or medical provider, medical supplies or services, debilitating illness or injury, health insurance costs, total medical expenses, receipt of medical reimbursements, visits to dentist while not covered by dental insurance, visits to health care provider while not covered by health insurance. Asked about respondent: Health conditions that prevent working, likelihood of working in the next year. [93-7, 96-3, 96-6, 96-9] Work Related Expenses, Child Support Paid and Child Care Poverty. Transportation to work, distance, commuter expenses, work-related expenses, cost / payment of childcare arrangements (including schooling), living situation of children, child support payments. [96-3, 96-6, 96-9] Work-Related Expenses. Asks about work-related expenses for each employer the respondent had during the reference period. Questions address various costs of working, such as union dues, licenses, special tools, and uniforms. Mode of transportation and mileage driven to and from work are determined, along with any parking or mass transit fees. (Also asked as Work-Related Expenses and Child Support Paid.) [84-5, 84-8, 854, 86-6, 87-3, 96-3, 96-6, 96-9, 96-12] Assets, Liabilities, and Eligibility. Collects information about the value of assets and debt on assets and expands on data gathered in the core questions. The intent of this topical module is to derive a comprehensive measure of household net worth and to collect information used to determine eligibility for federal assistance programs. To that end, the topical module includes selected additional questions needed to determine program eligibility. Some of the assets included are savings accounts, stocks, mutual funds, and bonds. Data on unsecured liabilities such as loans, credit cards, and medical bills are also gathered. Assets and liabilities that are held jointly are identified to prevent double-counting. The 1996 version of this module has seven sections: value of business; interes t earning accounts; stocks and mutual funds; mortgages; other assets; assets and liabilities; and real estate, shelter costs, dependent care, and vehicle ownership. (Also asked as Assets and Liabilities.) [84-4, 84-7, 85-3, 85-7, 86-4, 86-7, 87-4, 90-4, 91-7, 92-4, 93-7, 96-3, 96-6, 96-9, 96-12] Real Estate, Shelter Costs, Dependent Care and Vehicles. Identify owner of home, date purchased, details about debt on home, current property value, monthly cost of utilities, rent paid by household members, dependent care provided so household member could seek job / attend job training, ownership of other properties, ownership of vehicles, money owed on vehicles, ownership of boat / motorcycle / RV. Real Estate Property and Vehicles. Gathers information about housing tenure and financing, other real estate ownership, and automobile ownership. Home owners are asked a series of questions that allow the estimation of net real estate equity. Questions about vehicles address ownership, type of vehicle (i.e., car, truck, motorcycle), value, and amount owed. Those questions are also used in program eligibility simulations. (A variation of this module is asked as Real Estate, Shelter Costs, Dependent Care, and Vehicles.) [84-7, 853, 85-7, 86-4, 86-7, 87-4, 87-7, 88-4, 90-4, 90-7, 91-4, 91-7, 92-4, 92-7, 93-4, 93-7] Value of Business. Percent of business owned, value of business (before debts), value of debts owed against business. Interest Earning Accounts. Amount in jointly and singly held accounts / assets. Rental Properties. Joint / sole ownership of rental property (type, location, value, debts owed. Stock and Mutual Fund Shares. Joint / sole ownership of mutual funds and stocks (market value, debt held against accounts. Mortgages. Joint / sole ownership of mortgage, principal owed. Other Financial Investments. Equity in other financial investments 106 Chapter 9: The Survey Of Income And Program Participation Abt Associates Inc. Exhibit 9.2 SIPP: Content of Topical Modules, by Wave Wave 4 Annual Income and Retirement Accounts (repeated in Wave 7). Obtains respondent estimates of calendar-year business income and respondents. personal retirement plans. The module asks about businesses owned by respondents, gross income and expenses to such businesses, net income to such businesses, retirement accounts, including IRA, Keogh, and 401(k), and respondent participation in those retirement plans. [84-9, 85-5, 85-8, 86-5, 87-5, 88-5, 90-5, 90-8, 91-5, 91-8, 92-5, 92-8 93-5, 93-8, 96-4, 96-7, 96-10] Taxes (repeated in Wave 7). Includes questions about exemptions, calendar-year wages and salaries, income from businesses, itemized deductions, and earned income credits. Respondents are asked about federal and state income tax liabilities, exemptions, amounts owed for federal and property taxes, and amounts from a variety of tax schedules. To help ensure accuracy, interviewers encourage respondents to refer to income tax returns and other records. Historically, this module has been administered at least twice per panel, generally in the spring when respondents were likely to be preparing their tax returns for the prior year. (Also asked as Earnings and Benefits, and Property Income and Taxes.) [84-6, 84-9, 85-5, 85-8, 86-5, 87-5, 88-5, 90-5, 90-8, 91-5, 91-8, 92-5, 92-8, 93-5, 93-8, 96-4, 96-7, 96-10] Work Schedule. Collects information about the number of hours and days worked during a typical week in the fourth reference month. Questions about whether or not the respondent worked only at home on any days are included. [87-6, 88-3, 88-6, 89-3, 90-3, 91-3, 92-6, 92-9, 93-3, 93-6, 93-9, 96-4, 96-10] Child Care. Collects information about all child care arrangements, for all children under age 15, from mothers, single fathers, or guardians, regardless of labor force status. Those with children under age 15 are asked about the type of child care arrangements, who provides the care, the number of hours of care per week, where the care is provided, and the cost of the care. The module asks whether a relative or nonrelative cared for the child, and if the child was in school. Before the 1993 Panel, the module collected information about only one to two child care arrangements from mothers, single fathers, or guardians who were either working, in school, or looking for a job during the 4-month reference period. [84-5, 85-6, 86-3, 86-6, 87-3, 876, 88-3, 88-6, 89-3, 90-3, 91-3, 92-6, 92-9, 93-3, 93-6, 96-4, 96-10] Wave 5 School Enrollment and Financing. Seeks information about basic educational attainment, enrollment in public and private schools, and whether those in government programs differ from others in terms of financing their education and their sources of educational assistance. Asked of people aged 15 and older, the module includes questions to pinpoint the grade level of people enrolled in a general, technical, or business school; their pattern of full- or part-time enrollment; amount of tuition and fees; cos ts of room and board; and books and supplies. Specific sources of educational assistance, such as the GI Bill or employer assistance, are also determined. (Also asked as Education Financing and Enrollment.) [84-9, 85-5, 85-8, 86-5, 87-5, 88-5, 90-5, 90-8, 91-5, 91-8, 92-5, 92-8, 93-5, 93-8, 96-5] Child Support Agreements (Repeated in Wave 8). Helps determine whether money received as child support affects participation in government programs and whether lack of support from one parent causes the other parent to need government assistance. The module collects information about characteristics of child support agreements, the annual amount and frequency of payments, and provisions for health care costs. Additional questions cover custodial arrangements, contact with public agencies for assistance in collection of child support, frequency of contact with the absent parent, current place of residence of the absent parent, and reasons for nonaward of child support. Questions about paternity establishment status are also asked about children of women with nonwritten agreements and all never married women. [85-6, 86-3, 86-6, 87-3, 87-6, 88-3, 88-6, 89-3, 90-3, 90-6, 91-3, 92-6, 92-9, 93-3, 93-6, 93-9, 96-5, 96-11] Abt Associates Inc. Chapter 9: The Survey Of Income And Program Participation 107 Exhibit 9.2 SIPP: Content of Topical Modules, by Wave Support for Non-Household Members (Repeated in Wave 8). Provides information about respondents; routine payments supporting people who are not current household members. Includes both child support payments for own children under 21 years of age and payments made to (or for) people who are not children of the respondents, for example, an elderly parent in a nursing home or an adult child living away from home and in an entry-level job. Questions about child support include number of children supported, type and year of agreement, annual amount and method of payment, health care provisions and custodial arrangements, and amount of contact with the absent children. Questions about support for other persons outside the household include their relationship to the respondent, living arrangement, and annual amount of support paid. [84-5, 84- 8, 85-4, 85-6, 86-3, 86-6, 87-3, 87-6, 88-3, 88-6, 89-3, 90-3, 90-6, 91-3, 92-6, 92-9, 93-3, 93-6, 93-9, 96-5] Functional Limitations and Disability – Adults and Children (Repeated in Wave 8). Provides data that can be used to evaluate links between types of disability, the family financial situation, and program participation. This module is asked in three variations: overall, adult, and children. Adults are asked the standard Activities of Daily Living (ADL) and Instrumental Activities of Daily Living (IADL) battery of questions. Questions address physical and mental conditions affecting the respondent, the use of mobility aids, vision and hearing impairments, speech difficulties, lifting and aerobic difficulties, and the ability to function independently within the home. For those under age 22, the questions are modified, referring to ageappropriate activities (e.g., questions about work activities are recast to ask about analogous school activities). Questions about children also address the use of special education services. For those under age 15, the interviewer asks the questions of the designated parent or guardian. [90-3, 90-6, 91-3, 92-6, 93-3 for overall module; 92-9, 93-6, 96-5, 96-11 for separate children and adults modules] Employer-Provided Health Benefits. Collects data on the availability of health care benefits from employers and the demographics of workers with and without employer-provided health coverage. The module asks whether the plan restricts the respondent to specified doctors, if family members are covered, and whether any family members have pre-existing conditions not covered by the plan. The module also asks about longterm health care options. [96-5] Wave 7 Retirement Expectations and Pension Plan Coverage. Obtains information about the respondent's pension plan coverage for the most important current job or business, and information from persons currently receiving retirement benefits from a former job or business. Respondents are asked about their coverage and vesting in pension plans, types of plans, the reasons they are not included by or do not participate in plans, current contributions and amounts of money in their accounts if applicable, and how the money in their own plans is invested. Other questions concern loans from pension accounts and treatment of lump sums received from prior job pension plans. Respondents currently receiving pension income are asked about the types of pension they receive, provisions for cost-of-living adjustments, and health benefits. Respondents are also asked Industry and Occupation data about the job or business from which their pensions are received. (Also asked as Pension Plan Coverage [84-7].) [84-4, 85-7, 86-4, 86-7, 87-4, 90-4, 91- 7, 92-4, 93-9, 96-7] Home Health Care. Asks about the type and sources of help given to respondents who needed help with their personal care, household activities, and basic errands because of a health condition. Respondents are asked if caregivers were relatives or nonrelatives, and whether or not the caregivers were household members. This module also asks about members of the household who might have given such care, on a nonprofessional level, to a person outside the household. Questions determine the relationship of the caregiver and recipient(s) and the kind of care given. [88-6, 89-3] Children’s Well-Being. Asks the designated parent or guardian about the health of children in the household, care of the child by nonfamily members, activities the family does with the children (such as reading and outings), lessons and activities outside of school, rules for children's TV viewing, and the respondent's opinion about the quality of the neighborhood. The module obtains information about children in three age groups: under 6 years old, ages 6-11, and ages 12-17, for as many as seven children in each category. Certain questions target fathers or stepfathers who are not designated parents; other questions address whether the child attends a public or private school. Content of this module varies across different panels and waves; analysts should check the documentation for exact content. [92-9, 93-6, 93-9, 96-6, 96-11] 108 Chapter 9: The Survey Of Income And Program Participation Abt Associates Inc. Exhibit 9.2 SIPP: Content of Topical Modules, by Wave Wave 8 Adult Well-Being. Asks the reference person about consumer durables, living conditions, crime, neighborhood conditions, community services, basic needs, and food adequacy. This topical module assesses the standard of living of SIPP respondents. It is similar to Extended Measures of Well-Being and incorporates Basic Needs information that was asked as a separate module in 93-9. [93-9, 96-8] Abt Associates Inc. Chapter 9: The Survey Of Income And Program Participation 109 Chapter Ten: The Survey Of Program Dynamics The Survey of Program Dynamics (SPD) is a longitudinal survey initiated in 1997 that collects data on the economic, household and social characteristics of a nationally representative sample of the US population over time. It was created in response to PRWORA, in which Congress required the Census Bureau to continue collecting data on the 1992/1993 panels of the Survey of Income and Program Participation (SIPP). The primary goals of the SPD are to provide information on spells of actual and potential program participation over a ten-year period; to examine the causes of program participation and its long-term changes that result from implementing welfare reform; and to assess the effects of national welfare reforms, how these reforms interact with each other, and how they interact with employment, income and family circumstances. It is sponsored and administered by the U.S. Census Bureau. Survey Design Conceptually, the SPD universe includes all non-institutionalized residents of the United States. In practice, the sample was drawn from SIPP participants who were interviewed in the first wave of the 1992 and 1993 panels and were still being interviewed at the end of their panel. There were 34,609 households in this group. The 1992/1993 SIPP was a multistage, stratified sample of the US civilian non-institutionalized population. In preparation for the SIPP redesign described in the previous chapter, the Census Bureau cancele d the 1994 and 1995 panels. The 1992 SIPP had 10 waves of data collection, with the final interviews occurring in February to April 1995. The 1993 SIPP had nine waves, with the final interviews occurring in October 1995 to January 1996. The 1997 SPD Bridge Survey re-contacted households that were interviewed in the 1992/1993 SIPP panels, covering the gap in data collection between the close of the SIPP panels and the start of the SPD proper. Budget constraints necessitated a reduction in sample size of about half between the Bridge Survey and the 1998 SPD. The sample for the latter was a stratified random subsample of households interviewed in the Bridge Survey, comprising: • • • • • All households under 150 percent of poverty (based on the income of the primary family or primary individual) All households between 150 percent and 200 percent of poverty that contained children under age 18 Ninety percent of the households above 200 percent of poverty that contained children under age 18 Eighty percent of the households between 150 percent and 200 percent of the poverty threshold with no children under age 18 27 percent of households “in the balance”—childless households above 200 percent of poverty Abt Associates Inc. Chapter 10: The Survey Of Program Dynamics 111 It also included 27 percent of the SIPP cases that were institutionalized during the SPD Bridge and therefore not interviewed at that time. Some of these households were reintroduced to the sample starting in 2000. Further SPD interviews were conducted annually, in 1999, 2000, 2001, and 2002. Sample size: Linking together the 1992/1993 SIPP, the SPD Bridge, and the subsequent annual SPD interviewing waves, the sample sizes are as shown below: 1992/1993 SIPP: 1997 SPD: 1998 SPD: 1999 SPD: 2000 SPD: 2001 SPD: 35,291 30,125 16,395 16,659 18,716 (base sample + 3,456 households selected for 1997 SPD) 22,340 (base sample + 3,616 households selected for 1997 SPD but not interviewed, plus 5,540 households selected for the 92/93 SIPP but not interviewed) Methods of data collection: Data were collected in person. While the 1992 and 1993 SIPP panels used paper-and-pencil instruments, the SPD used CAPI. In addition, the self-administered adolescent questionnaire utilized an audiocassette-administered instrument (A-CASI). Response rates: The response rate for the 1992/1993 SIPP was 73.4 percent. The rates for the Bridge Survey and subsequent waves of SPD data collection through 2001 were 58.7 percent, 50.0 percent, 50.2 percent, 55.7 percent, and 63 percent. There has been particular concern about the higher attrition rates for low-income households. Survey Content The 1992/1993 SIPP collected a plethora of information on sampled households. Data collection occurred at intervals of 4 months. At each contact, the household roster was updated and detailed data were obtained on household members’ employment and earnings, unearned income sources, and participation in means-tested programs. Occasional topical modules for these two panels included the following: • • • • • • • • • • • • 112 Children’s well-being Education and training history Employment history Extended measures of well-being Fertility history Functional limitations and disabilities Health status and health care utilization Household relationships Marital history Medical expenditures and work disability Migration history Real estate, shelter costs, dependent care, and vehicles Chapter 10: The Survey Of Program Dynamics Abt Associates Inc. • • • • • • • • Recipiency history Retirement expectations and pension plans School enrollment and financing Selected financial assets Support for non-household members Taxes Work disability history Work schedule The Bridge Survey used a modified version of the March 1997 CPS Demographic Supplement to collect information on households’ resources and economic circumstances in 1996, including total family income, work experience, earned income, programmatic income, other unearned income, health insurance, non-cash program participation, and migration. Some items also were collected for 1995 pertaining to receipt of cash assistance, food stamps, and Medicaid. The annual SPD survey included sections on household composit ion, employment and earnings, income sources and amounts, eligibility and assets, vehicle operating expenses, use of public transportation, educational enrollment, work training, substance abuse, functional limitation and disability, health care utilization and insurance, food expenditures and food security, children’s school enrollment, enrichment activities, disability, and health care utilization, mother’s work schedule, child care, child support agreements, contact with absent parent, child well-being, child’s positive behaviors and social competence, family routines, conflict between parents, marital relationship and conflict, and parental depression scale. Suitability for Research Outcome measures: Since the SPD incorporates the measures collected as part of the initial SIPP interviews, most of the description of the SIPP in the previous chapter is relevant for describing the measures available in the SPD as well. Like the SIPP, the SPD includes marital and cohabitational status, and a marital history module. Changes in household composition between waves can be used to construct an approximate cohabitational history, limited to the length of the panel. This information is only available on an annual basis, however. Unlike the SIPP, the SPD collects measures about union quality for couples with children, on an annual basis. Measures of demographic influences: The SPD is an appropriate data source for investigating key demographic influences on union behaviors: teen and non-marital childbearing, transitions to parenthood, and effects of cohabitation on marriage. Like the SIPP, the SPD includes complete marital and fertility histories, and reports the relationships among all members of the household. The marital and fertility histories permit the study of the influences of both early and non-marital childbearing on union formation and dissolution. The extension of the longitudinal data collection period increases the usefulness of this survey for studying any long-term influences as compared to the SIPP. Relevant covariates are measured prospectively throughout the 10-year period of data collection, allowing adequate specification of correctly ordered causal models. Some aspects of the influence of the transition to parenthood can be also examined, through the use of fertility histories. For individuals experiencing births during the survey period, this line of research can be bolstered by the availability of some relevant measures of positive and negative interactions in relationships. The Abt Associates Inc. Chapter 10: The Survey Of Program Dynamics 113 influence of cohabitation on marriage can be examined only for union transitions occurring during the 10-year period of data collection. Measures of economic influences: The SPD is an appropriate data source for investigating key economic issues relevant to union behaviors. It has extensive and highly detailed measures of both male and female labor market experiences, especially for low-income populations. Information is available annually on earned and unearned income, program participation, work training, disabilities, and mother’s work schedule. Labor force measures include weeks worked, hours worked per week, wage/salary, fringe benefits, type of employer, occupation, work activities and duties. Topical modules from the SIPP provide further information on education and training, migration, and recipiency histories. The recipiency history and detailed information about program participation and eligibility make the SPD particularly relevant to studying disadvantaged populations. While the job and recipiency histories could be linked to the marital histories, the SPD is more appropriate for examining causal relationships between more current economic influences and prospective measures of union transitions during the survey period. Measures of socio-cultural influences: The SPD does not collect data on respondents’ views on marriage and cohabitation or men’s and women’s roles, so that the influence of these factors on union behaviors cannot be examined directly. However, indirect analysis of these types of influences is possible by capitalizing on the measures of male and female employment and earnings available in the SPD, as well as the measures of fertility, household structure, child care and family routines (eating breakfast together, eating dinner together, serving dinner at a regular time). These measures offer evidence of men and women’s breadwinning and household labor behaviors, and can be linked with union behaviors to offer indirect evidence of the influence of gender roles on union outcomes. Measures of psychological influences: The SPD includes limited measures of psychological influences relevant to union behaviors. The parental depression scale and the measures of substance use could be examined as psychological influences on the measures of union quality and conflict. Prospective measures of these psychological influences can be combined with information in union quality or changes in union status measured at a later wave of the survey to develop causal models with correct temporal ordering. Special strengths: The design of the SPD is superlative. All types of households and non- institutionalized individuals are represented. Extraordinarily rich data have been collected on sample members over a period of ten years, including retrospective histories of employment, marriage, migration, and welfare receipt, and current information on union quality, work schedule, substance abuse, and mental health. The multiple observations would support estimation of fixed effect and linear growth models. The relatively large sample size permits focus on detailed subgroups, which is particularly useful for extending research around issues of disadvantage. The rich array of measures around individual and household economic well-being in the SPD could be capitalized upon to investigate differing approaches to defining and describing disadvantaged populations and their union patterns. The potential for studying union behaviors among disadvantaged populations is further strengthened by the SPD’s emphasis on program participation. Limitations: A clear and very serious limitation of the SPD, however, is the low response rate. User Notes for the second (interim) SPD Longitudinal File include this caution: 114 Chapter 10: The Survey Of Program Dynamics Abt Associates Inc. The original SPD sample is five to six years old and has undergone a substantial amount of attrition (sample loss) from 1992 (or 1993) through 1998. As a result, estimates from this file are not as representative of the U.S. population as a cross-sectional survey or a longitudinal survey with less sample attrition would be. Therefore, results should be viewed with caution. (www.sipp.census.gov/spd/long/usernote_2lgt.htm.) In addition, it should be noted that this file contains data collected using three different survey collection vehicles: the 1992/93 SIPP paper instruments that were used to produce data for calendar years (CY) 1992, 1993, and 1994, a modified March CPS CAPI instrument that was used to collect data for calendar year 1996, and the 1998 CAPI instrument that was used to collect data for calendar year 1997. Therefore, the different questions and modes of interview used to produce the estimates should be considered when analyzing changes over time. Using the Data The SPD website is http://www.sipp.census.gov/spd/spdmain.htm. Documentation: Two key methodological reports for users are: Survey of Program Dynamics, User’s Guide (2002). U.S. Department of Commerce, Economics and Statistics Administration, U.S. Census Bureau. Survey of Program Dynamics (SPD), First Longitudinal File , U.S. Census Bureau, 2001. These are available online at the SPD website. Data Availability: Public use data files currently available from the Census Bureau include: • • • the unedited 1998 calendar year file, a file based on the 1998 SPD (covering 1997) six cross-sectional files, for calendar years 1997 through 2002 three longitudinal files covering 1992 through 1997, 1999, and 2001, respectively (excluding 1995). Files are available for purchase from the Marketing Services Office, Customer Services Center, US Census Bureau, Washington DC, 20233. These files are on CD-ROM, in ASCII format. Alternatively, data extracts may be downloaded from the SPD website via FERRET. Available files on FERRET are the Bridge survey and a longitudinal file covering 1992 through 1999. Note that the public use files do not include the Self-Administered Questionnaire (SAQ) data, which must be specially requested from the Census Bureau Research Data Center. This segment includes the items on marital relationship and conflic t, and parental depression. Reports and Articles: A list of documentation and working papers is provided at http://www.sipp.census.gov/spd/pubsmain.htm. Available working papers include, for example, Abt Associates Inc. Chapter 10: The Survey Of Program Dynamics 115 Bass, L. E. & Downs, B. (1999). What can the SPD Adolescent SAQ tell us about the wellbeing of adolescents in the aftermath of the 1996 Welfare Reform Act? SPD Working Paper No. 99-4. Hofferth, S. L. (2002). Use of the Survey of Program Dynamics for studying the implications of welfare reform for familie s and children. SPD Working Paper No. 2002-1. Data Elements The contents of the Bridge Survey and the annual SPD Survey are listed in Exhibits 10.1 and 10.2, respectively. 116 Chapter 10: The Survey Of Program Dynamics Abt Associates Inc. Exhibit 10.1 SPD: Content of 1997 SPD Bridge Survey Hispanic: As of November 19, 1996: reference person in residence, other household members in residence. Asian Pacific Islanders: Specify Asian or Pacific Islander group of origin. Social Security Number Family Income Work Experience in 1996 Work at business / job, temporary/part-time/seasonal, time on layoff/job searching (# weeks), reason for not working, reasons for not looking for a job, number of weeks / months working, number of employers, hours/week. Weeks worked 35 hours or less: number of weeks, reason. Name of longest job, class of worker, name of employer/business, type of industry, type of work, most important duties, type of employer (government, private company, non-profit, etc.), business owner, number of other employees. Earned Income in 1996 Income earned from employer, number of pay periods, tips / bonuses / overtime pay / commissions, net earnings / loss from business/farm, money earned from other employers. Unemployment and Workers Compensation in 1996 Receipt of state or federal unemployment compensation, Supplemental unemployment benefits, union unemployment or strike benefits, worker’s compensation: amount received, number of payments. Social Security in 1996 Social Security (SS) payments, family members receiving SS, amount received, number of months, Medicare deduction, reasons for receiving SS. SS funds received on behalf of children less than 23 years of age, amount, number of months, reasons for receiving SS, specify children. Supplemental Security Income (SSI) in 1996 SSI payments, SSI for children: family member receiving SSI, amount received, number of months, changes in payments, reasons for receiving SSI. Public Assistance Public assistance, welfare payments, VA payments: family members receiving payments, type of payment (public assistance: AFDC, ADC, TANF), amount received, periodicity of payment, annual income questionnaire for VA payment, amount received, periodicity of payment. Survivor Benefits Survivor benefits (widow’s pension, estates, trusts, insurance annuities, other): family members receiving payments, source of income (company, government, US Railroad, Worker’s Compensation, Black Lung Survivor, regular payments from trust / estate / annuity), periodicity of payment, amount received. Disability Income Disability preventing/limiting type of work: family members, retire/leave job due to health reasons: family members, income as result of health problem, source of income, periodicity of payment, amount received. Retirement and Pension Receive pension or retirement income: family members, source of income, periodicity of funds, amount received; 1996 employers / union with pension/retirement plan for employees. Earned Interest Funds in money market, savings, bonds, treasury notes, IRAs, CDs, interest earning checking: family member (over age 15) receiving, periodicity of interest income, amount received. Dividends Stocks in corporations or mutual fund shares: family member receiving, dividend income, periodicity, amount. Property Income Rented property (land, business property, apartments, homes), income from roomers / boarders, income from royalties, income from estates / trusts: family member receiving funds, amount, periodicity. Educational Assistance School attendance beyond high school in 1996, educational assistance for tuition, fees, books or living expenses, family member receiving funds, type of assistance, amount, periodicity. Abt Associates Inc. Chapter 10: The Survey Of Program Dynamics 117 Exhibit 10.1 SPD: Content of 1997 SPD Bridge Survey Child Support and Alimony Receipt of child support payments, family member receiving funds, periodicity, amount, number of payments. Regular Financial Assistance Regular financial assistance from non-household friend/relative, family member receiving assistance, periodicity, amount, number of payments. Other Money Income Income from hobbies, home businesses, farms: family members receiving funds, periodicity, amount, number of payments. Income from severance pay, welfare, foster child care payments, other money income: family member receiving funds, periodicity, amount, number of payments. Health Insurance Health insurance through employer or union: policy holder, family members covered (in and outside the home), insurance premium; plan purchased directly: policy holders, family members covered (in and outside the home); plan owned by someone outside of the household: family members covered; Medicare: family members covered; Medicaid: state name, family members receiving Medicaid, duration; CHAMPUS/CHAMPVA/military health care/Indian health services: family members covered, respondent’s plan; coverage last week: family members with insurance last week, type of plan, overall health rating. School Lunches Children (aged 5 - 18) who usually eat school lunch, free/reduced price lunch. Public Housing Public housing through local housing authority or other public agency, government rent subsidy. Food Stamps Receipt of food stamps, family members covered, periodicity, amount received, number of months received. Energy Assistance Energy assistance from government, amount received. Migration Living in same home last year: name of city/town, state, zip code, county, country, number of housing units in structure. Questions about 1995: public assistance of welfare in 1995 (not including SSI or food stamps), family members receiving payments, AFDC/ADC, amount in public assistance, food stamps: monthly value, duration; Medicaid: family members covered, name of state program, duration. 118 Chapter 10: The Survey Of Program Dynamics Abt Associates Inc. Exhibit 10.2 SPD: Content of 1999 SPD General Reference Section Household composition / roster (name, relationship to reference person), address, telephone number. Front Section Verify address, mover situations. Race, gender, number of household members, status of living quarters (owned/rented/occupied without payment of cash rent, person who rents / owns); those who left household: new address of former household members, reason for leaving, date of departure. New address, if whole household has moved, type of housing unit, group quarters unit, number of housing units in structure, public housing; rent subsidy: months received; new additions to household: name, usual residence, date residence begun, reason for joining household, gender; household members: relation to reference person, gender, children (step/adopted/foster), birth date, age, marital status, past experience being widowed or divorced, participation in US Armed Forces: date of active duty, highest level of education, GED, race / origin of descent, birth country, citizenship status, language spoke at home, English proficiency, time living away from household, SSN / RRN. Employment and Earnings Employment during reference year, temporary/part-time/seasonal, layoff (date of layoff / return to work), time spent looking for work (date), reason for not looking for job, weeks worked, vacation/sick leave, employers (number, name, type, family farm/business, address, job, start/end dates, reason for leaving, duties, number of other employees), hours per week, unemployment benefits. Earnings: amount earned in 1998 before deductions (from job, from business/farm), tips/bonuses/overtime/commission. Benefits: pension/retirement plan, health insurance coverage, paid vacation, paid sick leave, tuition assistance. Current employment status. Income Sources Total combined household income, household members receiving unemployment (type) / Worker’s Compensation (source) / Social Security payments (for self / on behalf of child) / SSI (for self / on behalf of child), food stamps / welfare / AFDC or ADC / WIC / Emergency Assistance / General Assistance (for transportation, child care or other needs) / cash assistance (amount), free or reduced price school meals, energy assistance, Foster Child Care payments, Veteran’s payments (type), income as a survivor or widow from pensions / estates/trusts/annuities or other survivor benefits (source). Physical /mental / or other problem preventing work / limiting type of work, possibility of working in 12 months, retire/leave job due to health reasons, disability income (source), pension / other retirement income (source), money in savings / interest-earning checking / money market fund / bonds / treasury notes / CDs, mutual fund shares / 401k / IRA mutual funds, stocks, rental property, rental income from roomers or boarders, royalties, estates/trusts, alimony or maintenance payments, child support through parent or welfare/child support agency, financial assistance from friends or relatives not in household, other sources of income (National Guard or Reserve pay, casual earnings, farm income, lump sum payment, assistance from charitable group). Independent/Dependent Comparison Comparisons between 1997 and 1998 on receipt of Unemployment Compensation, Worker’s Compensation, State Worker’s Compensation / employer’s insurance / own insurance / other source, Social Security paym ents, SSI, public assistance payment, veteran’s payments, survivor payments, disability benefits, retirement benefits, alimony, child support payments. Amounts Income and benefits received in 1998 from previously identified sources: periodicity, amount, weeks funds received, reasons for applying for benefits / reasons benefits ended, consequences of lost benefits. Eligibility and Assets Home: owner / purchaser, market value, mortgages, home equity loans, monthly mortgage payments, total property taxes, annual insurance premium, monthly home equity loan payment, monthly costs: rent, electricity, heat, basic phone service, water, sewer, persons paying mortgage/rent and utilities, remaining principal on mortgage. Cars/trucks owned: number, make/ year/model, money owed, primary use. Other debts (credit cards, student loans, medical or legal bills, loans from relatives), total amount of debts. Pay child support: amount, number of children. Alimony: amount. Abt Associates Inc. Chapter 10: The Survey Of Program Dynamics 119 Exhibit 10.2 SPD: Content of 1999 SPD Vehicle Operating Expenses Public transportation in area: use. Transportation to work (type, cost, problems related to work/job training or school). Educational Enrollment Enrolled in school (full / part time): high school/college/vocational/other, grade or level, reasons for enrolling (wanted to, helped meet requirement, both), financial aid (type), costs. Work Training Job training: provided / referred by social services, reasons for enrolling (wanted to, helped meet requirement, both), type, duration (expected and actual); job search help / job placement: provided / referred by social services, reasons for enrolling (wanted to, helped meet requirement, both), community service/tryout employment/workfare. Substance Abuse Alcohol consumption: number of drinks per day, interference with school/work/home life (number of times), potential for injury under the influence, related emotional psychological problems, urge to drink, time drinking, time spent getting over the effects of alcohol, drink more/longer than intended. Drug abuse (prescriptions included): nerve pills, stimulants, prescription painkillers, inhalants, marijuana, cocaine/crack/free base, LSD or other hallucinogens, heroin, interference with school/work/home life (number of times), potential for injury using drugs, related emotional psychological problems, urge to use, time using, time spent getting over the effects of drugs, use more/longer than intended (number of times), need to increase use to elicit same effect. Functional Limitation and Disability Overall health rating, limitations to school work due to health condition, special education, vision impairment, special aids (i.e., wheelchair, hearing aid, hearing impairment, difficulty lifting, difficulty walking, need help with daily routine (i.e., dressing, bathing, chores). Health Care Utilization Overnight hospital stay / admittance to psychiatric unit: number of times, reason, number of nights; number of doctor / dentist visits, place to seek medical advice (i.e., clinic, hospital, doctor’s office, etc.); medical expenses: out-of-pocket expenses. Health Insurance Type of coverage (employer-provided, union provided, Medicaid, military health, Indian Health services, state plan, other government program): policyholder, coverage, costs, months covered, current coverage; reasons for not having health care coverage. Health Care Utilization While Uninsured While not insured: doctor visits, treatment for illness/injury, preventive care, treatment drug/alcohol problem, type of health care services, costs. Food Expenditures Expenditures: cost of groceries / non-food items / eating out per week. Food Security Overall food security, reasons for food shortage, food security assessment (series of questions). Child-Related Questions Children’s School Enrollment Head s tart / grade level, months enrolled, costs, weeks attended, public/private, gifted/advanced program, grades repeated, suspended/excluded/expelled, change of schools, school expenditures. Enrichment Activities Sports team participation, lessons (i.e., mus ic, dance, computer), clubs/organizations, TV rules / TV watching habits, reading to child, outings with child, attendance at religious service / event, gang participation, arrests, job. 120 Chapter 10: The Survey Of Program Dynamics Abt Associates Inc. Exhibit 10.2 SPD: Content of 1999 SPD Children’s Disability Overall health rating, developmental/learning disability, difficulty with age appropriate activities due to health, limitations to school work due to health, special education, emotional/behavioral problem, vision impairment, use of special aid (i.e., wheelchair, cane), hearing impairment. Children’s Health Care Utilization Overnight hospital stay: number of times, reason, number of nights; number of doctor / dentist visits, place to seek medical advice (i.e., clinic, hospital, doctor’s office, etc.); medical expenses: out-of-pocket expenses. Mother’s Work Schedule Work status of mother during year, part-time/evening/weekend work, weekly schedule, job search / training activities, school attendance. Child Care Type of child care (family member, non-relative friend, child care center/day care/pre-school, head start, other): months care provided, hours per week, reasons for child care, cost, assistance with cost of care, reasons for not receiving financial assistance, difficulty working related to child care. Child Support Agreement Parent living outside of home (reason), legal responsibility for child, arrangements (legal or otherwise) regarding financial support of child, 2+ children covered by same agreement, amount of financial support (supposed and actual), payments sent directly from parent / court / welfare or child support agency / etc., ever sought help from agency in obtaining support funds, custodial arrangement for child, other types of child support (i.e., health insurance, housing, birthday or other gifts, food, child care or tuition). Contact with Absent Parent Date of last contact with absent parent, live in same county/town and city as absent parent, periodicity of phone conversations / written correspondence / visits / overnight visits, conflict between parents over child’s home, parenting, money / child support, time spent with child, physical conflict. Extended Measures of Child Well-Being How often do you do the following with child: praise child, talk/play one-on-one, laugh, do something special, look at picture books, having an outing, play sport / game, read, tell stories, do arts/crafts, household chores, talk about nature/science, visit library / museum, attend an overnight retreat; how often does child: read, visit library / museum; child’s progress in school, child’s feelings towards school / school work / homework, routine bedtime, regular breakfast time. Behavior / emotional problems: sleep, speech, sadness/depression, temper tantrums, nervous, concentration, trouble getting along with other kids, inferior, acts young for age, lying / cheating. Positive Behaviors / Social Competence Child is/does: warm and loving, gets along with other children, admired/liked by other children, shows concern for feelings of others, helpful/cooperative, considerate/thoughtful, share, plan ahead, try to make family proud, express thoughts/feelings, help others. Family Routines Periodicity with which: family eats breakfast together, eats dinner together, dinner is served at a regular time; home computer. Conflict Between Parents (self-administered) Periodicity of disagreements about: household tasks, money, spending time together, sex, in-laws, children. Periodicity of family members: fighting, losing tempers, getting angry and throwing things, calmly discussing problems, criticizing each other, hitting each other out of anger. Child bothers parent, parent gives up life to meet child’s needs, anger with child, trapped by parental responsibilities. Marital Relationship and Conflict (self-administered) Overall happiness in relationship, possibility of separation, physical arguments. Parental Depression Scale (self-administered) How often (in past 30 days) did you feel: sad, nervous, restless, hopeless, that everything was an effort, worthless. Interference of feelings with life and activities. Abt Associates Inc. Chapter 10: The Survey Of Program Dynamics 121 Exhibit 10.3 SPD: Conflict between Parents How often in the past year have you and your (spouse/unmarried partner) had disagreements about each of the following? [“never” to “almost every day”]: • Household tasks • Money • Spending time together • Sex • In-laws • The children Next are some statements about how families get along and settle arguments. For each one please indicate how strongly you agree or disagree that the statement is like the family you currently live with. [“strongly agree” to “strongly disagree”]: • We fight a lot in our family. • Family members lose their tempers. • Family members sometimes get so angry they throw things. • Family members always calmly discuss problems. • Family members often criticize each other. • Family members sometimes hit each other in anger. Please tell me how often the following statements are true for you [“all of the time” to “none of the time”]: • My (child/children) (does/do) things that really bother me a lot. • I find myself giving up more of my life to meet my (child’s/children’s) needs than I ever expected. • I feel angry with my (child/children). • I feel trapped by my responsibilities as my (child/children’s) parent. • I feel my (child/children) is/are much harder to care for than most children. Self administered: • Taking all things together, how happy are you with your relationship with your (spouse/partner)? [“completely happy” to “not too happy”] • How often have you and your (spouse/partner) discussed or considered separating during the past few months? [“often” to “never”] • Sometimes arguments between partners become physical. During the last year, has this happened in arguments between you and your(spouse/partner)? 122 Chapter 10: The Survey Of Program Dynamics Abt Associates Inc. Chapter Eleven: Research Opportunities In this final chapter, we consider the extent to which the characteristics of the data sources described here fit to the research needs discussed in the companion report to this one, Fein et al. (2003). That report focuses on 10 themes in the literature explaining the determinants of marriage and cohabitation among disadvantaged Americans. Although not exhaustive, our ten broad influences cover much of the important basic research on union formation, quality and dissolution. The practical implications of the data needs associated with each of these themes are discussed in turn below. In the discussion of each theme, the characterization of research issues and data needs is drawn from the companion report. We then assess the nine data sources, pointing out their main strengths and weaknesses for addressing the key research issues. Exhibits in Appendix A summarize the salient characteristics of the nine data sets with respect to each theme. Theme 1: Teen and Non-Marital Childbearing Research issues: To what degree might reducing early and non-marital childbearing lead to more marriages? Researchers have used event history analysis to trace the connections between nonmarital fertility and subsequent union experiences. Teens and adults who do not marry within a short period after birth face substantially diminished long-term marriage prospects. The near-term aftermath of a premarital pregnancy is a somewhat elevated likelihood of marriage, although legitimization rates for non-marital pregnancies (i.e., “shotgun marriages”) have greatly diminished in recent decades, especially among blacks. For single men and women, cohabitation has become nearly as likely a response to pregnancy as marriage. Both pregnancies and births lead to sharp increases in marriage among cohabitors, however. Although shotgun marriages are unstable compared to marriages formed prior to the onset of childbearing, there does not appear to be a greater risk of marital disruption for women who marry some time after a non-marital birth (most likely to someone other than the father). The effects of teen childbearing on later marital stability have not been ascertained, although there is strong evidence of greater instability for teen marriages. The impacts on subsequent union quality of early or non-marital childbearing also are largely unexplored. Future research on this topic should put more emphasis on understanding why non-marital childbearing reduces marriage prospects and the circumstances that may moderate its effects. Indications that nonmarital births are premised on weak marriage expectations from the start suggest a need for further research on the determinants of non-marital childbearing. Data needs: Addressing these questions requires complete and detailed fertility, marriage, and cohabitation histories, and measures of union quality at several points in time. It is necessary to distinguish between biological and stepchildren of couples. A wide range of potential explanatory measures are needed to analyze reasons for linkages between non-marital childbearing and union outcomes. Selection bias can perhaps be addressed by using siblings and women who miscarry as comparison groups. Assessment of data sources: With the exception of NSAF, which is cross-sectional and does not collect marital histories, all of the nine data sets have some potential to address these questions. CPS is relatively weak, however, because it is strictly retrospective; no measures are collected at the time Abt Associates Inc. Chapter 11: Research Opportunities 123 that each child is born. NSFG is also limited to retrospective measures, but it collects more detailed information about conditions at the time of the child’s birth than do some other surveys. A particular strength of the NSFG is that it offers what is likely the most complete (albeit still underrepresented) measures of abortion and miscarriages; this could be helpful for addressing concerns about selection bias, and better understanding links between fertility and union decisions. Although SIPP identifies children born during the course of the panel, the follow-up for these births is probably too short to be particularly useful for identifying anything but immediate influences on union behaviors. Fragile Families has both strengths and disadvantages for this area of research. It allows tracing forward for up to three years the consequences of a birth on union formation, quality and stability. Although this short follow-up period is a distinct disadvantage for researchers interested in understanding longer term influences, Fragile Families is an excellent and detailed data source for investigating more immediate responses to a teen or non-marital birth. In addition, it collects good attitudinal information from both partners. Comparison groups could be constructed based on marital status and age of mother, although the latter comparisons would run the risk of confounding mother’s age at time of birth with cohort effects. NLSY and PSID are both attractive because they are prospective and ongoing for several decades: sample members and their unions are directly observed before and long after any non-marital births. The key difference is that the PSID represents a full cross-section of the population with regard to age, while the NLSY is a single cohort. NSFH also has good potential because it spans 14 years and includes full information on marriage, cohabitation, and fertility. Finally, the SPD allows tracing of consequences over the 10-year course of the panel. In addition to standard information on demographics, household composition, and economic resources, these four data sets have rich contextual data of varying types. NSFH collects information from both partners on attitudes towards marriage. The NLSY, PSID, and SPD each have an array of occasional topical modules in addition to their comparable core data collected annually. The NLSY has included measures of self-esteem, time use, and vocational aptitude; the PSID has measured achievement motivation and time use; the SIPP modules available to the SPD include extended measures of well-being and recipiency history. Hence a choice among these four data sources depends on the researcher’s particular topical interests. Theme 2: Transition to Parenthood Research issues: In the period following first births, some couples may benefit from services that help them to adjust better to their new roles as parents and stresses that can affect the quality of their relationships. Careful longitudinal observation of convenience samples of middle -class white married couples document consistent, albeit modest, declines in average relationship quality in the initial year or two after a first birth. Larger declines have been documented for couples where there was poor communication, depression, and stress prior to the birth and when births resulted from unplanned pregnancies or were girls. Researchers have not made much progress in distinguishing between the direct effects of births and other factors that lead to declines in marital quality over time. The causal problem is complicated by the fact that children also bring couples new pleasures, and, on average, tend to increase union stability. Although the ratio of stress to satisfaction may be higher for poor couples and those in informal relationships (who typically are less committed to each other), there has been very little research on transitions to parenthood among these populations. A broader research agenda on the effects of children on unions also would focus on new stresses that arise when children reach adolescence and the complexities arising in step-parenting situations. 124 Chapter 11: Research Opportunities Abt Associates Inc. Data needs: The key data requirement for this research is measures of relationship quality from at least two points in time (before and after the focal birth), but preferably more. Data on parent-child relationships are also valuable, as are contextual variables on circumstances surrounding the pregnancy, economic stressors, and other life events. The sample should include some or all of the populations of special interest: single mothers, couples with stepchildren, cohabitors as well as married couples. Assessment of data sources: Plausible candidates, based on the availability of measures of union quality, are Fragile Families, NLSY, NSFH, and SPD. Each of these data sets has its pluses and minuses for this research. Fragile Families focuses on the birth of a child, and has good information about the couple’s relationship with each other and with the child. But it identifies the sample of couples with a new child retrospectively, hence does not measure quality of the relationship before the child’s birth. Also, it does not distinguish between the couple’s other stepchildren and biological children. The other three data sets allow prospective identification of couples having children. NLSY measures union quality annually, and has collected a plethora of data on sample members’ characteristics throughout their young adulthood. A drawback is the age of the sample —by 2000, the youngest members are already 36. NSFH is attractive because of the complete data that are collected from both members of the couple. The quality measures, however, are widely spaced, so that they cannot be closely linked to the birth of a child. The first follow-up occurred six or seven years after the baseline interview, and the next follow-up about eight years later. The SPD has many attractive features, including a nationally representative sample that covers all household types, rich historical data from the topical modules in the SIPP 1992 and 1993 Panels, and annual measures of union quality. The high rate of sample attrition is however a major shortcoming. Theme 3: Effects of Cohabitation on Marriage Research issues: There are concerns that cohabitation is replacing marriage with a less stable arrangement and that pre-marital cohabitation may engender attitudes detrimental to subsequent marital quality and stability. Descriptively it is clear that cohabitation has increased at the same time marriage was increasingly being postponed, but the degree to which cohabitation is a cause of later marriage has not been established. Analysis of the 1980s shows little evidence of increasing fertility rates within informal unions, although increased cohabitation meant a higher fraction of non-marital births were to cohabitors. Judging from actions following non-marital pregnancies, cohabitation is mainly a prelude to marriage for whites and an alternative to single parenthood for blacks. The most sophisticated studies have found that, although cohabitation engenders somewhat more liberal attitudes towards divorce, it does not increase the likelihood of marital disruption. Replication to confirm these results would be helpful. Indications of counterbalancing positive effects through increased age at marriage and improved information on potential spouses deserve further research. Initial findings of a negative association between cohabitation and marital quality also would benefit from re-analysis using more sophisticated techniques for addressing biases. The increased popularity of cohabitation suggests a need for ongoing study of recent cohorts to see if cohabitation is becoming a more fully institutionalized family form, as in parts of Europe. More research is needed on the decision processes leading from cohabitation to marriage. Research on cohabitation should take a Abt Associates Inc. Chapter 11: Research Opportunities 125 wider variety of forms of cohabitation as its object, rather than treating cohabitation as a uniform outcome. Data needs: Studying this topic requires histories of cohabitation and marriage for both men and women, with repeated measures of union quality. Sufficient samples at least through peoples’ 30s are needed to distinguish postponement from permanent avoidance of marriage. Comparisons of earlier and later cohorts are of great interest. Alternatively, it may be possible to use different crosssectional surveys to perform cross-cohort comparisons. Lichter and Graefe (2001), for example, used the 1987-88 NSFH and 1995 NSFG to compare the effects of nonmarital births on successive cohorts. Characteristics of the cohabiting relationship to be documented include regularity or frequency of cohabitation, interpersonal behaviors within unions, cohabitors’ relationships with family members and friends, and attitudes towards childrearing, commitment, and sexual exclusivity. Assessment of data sources: Three data sources—Fragile Families, NSFG Cycle 6, and NSFH— stand out as having good descriptions of cohabitational union characteristics, union quality, and sample members’ attitudes towards marriage. Overall, NSFH has the broadest applicability. It collects information over 14 years, and data collected on older focal children in later waves could be contrasted with analogous data collected on primary respondents in earlier waves. Although Fragile Families has excellent information about its sample, it is limited to a single cohort selected on the birth of a child. Thus it cannot be used to compare cohabitors that do and do not have children, or cohabitors in earlier or later periods. NSFG Cycle 6 is a single cross-section, with retrospective rather than prospective union histories. Hence the attitudinal information should be used for description rather than for causal inferences. Adding data from prior NSGF cycles would permit comparison of experiences across periods, with some ability to adjust for the confounding of age or cohort influences. Of the remaining data sources, the two long-running surveys, NLSY and PSID, would support estimation of duration of marital versus cohabitational unions, and the PSID would do so for different cohorts. Theme 4: The Surge of Low-Income Women Into Jobs Research issues: How has moving low-income women from welfare to work affected their chances for forming and maintaining healthy unions with men? The traditional view holds that women’s employment deters unions by reducing the advantages of role specialization within marriage, but more recent thinking has it that women’s earnings are increasingly necessary for marriage in an era of rising consumption standards. Analysts using different datasets, measures, and statistical techniques alternatively have found both positive and negative effects on marriage from women’s employment and earnings. Analyses of racial differences similarly have not found consistent patterns. The findings are difficult to weigh, given that they are subject to multiple sources of bias (e.g., selection, reverse causation, and restriction to young adult years) and cover a period during which causal effects may have been changing. More confirmation is needed of findings that effects of women’s economic status may be growing more positive, and more work is needed on the factors that may be contributing to such a change (e.g., increasing acceptance of female labor force participation, increasing consumption aspirations). Several studies suggest that the effects of women’s earnings on marriage may be positive at lower income levels and negligible or negative at higher income levels. It would be useful to examine interactions between women’s economic status and a wider array of indicators of socioeconomic disadvantage. Finally, more direct investigation of causal mechanisms would be useful—especially the degree to which women perceive greater bargaining power and 126 Chapter 11: Research Opportunities Abt Associates Inc. latitude to enter and leave relationships, whether men see women’s earnings as a positive attractor, and how work affects social networks and associated marriage and cohabitation possibilities for single mothers leaving welfare. Data needs: Research on this topic requires good measures of employment, including detailed descriptions of jobs. Data on spells of welfare receipt and more subjective aspects of economic status are also desirable, all preferably collected prospectively. Threats of selection bias can be addressed by exploiting variation in economic environments and incorporating rich measures of women’s abilities, attitudes, and temperaments. Interest in low-income women demands data on a sufficient sample size of disadvantaged women, however defined. Assessment of data sources: All of the nine data sets collect information on employment and earnings, including occupation and hours. None of them directly support a distinction between two ways in which women’s employment could increase union formation—through increasing her attractiveness as a partner versus increasing her access to potential partners. A distinction could perhaps be made indirectly by examining variations in women’s occupations and work schedules, which could be correlated with opportunities to meet men. Analyzing the potential effects of women’s employment on their union formation requires longitudinal data, and information on single women. The five data sets that meet these requirements are NLSY, NSFH, PSID, SIPP, and SPD. The wide spacing of the NSFH interviews renders it not very useful for this purpose, however. The NLSY and PSID—which have excellent measures of respondents’ background characteristics—have been used in this regard in the past (NLSY: Lichter et al. 1992; PSID: Hoffman, Duncan and Mincy 1991). An attractive feature of the SIPP is its subannual measures of employment. It is not as strong as the NLSY, however, with regard to measurement of potential confounders such as ability or attitudes. SIPP and SPD are both particularly relevant for examining relationships between employment—especially welfare-to-work transitions—and union behaviors for low-income women. NSAF offers some opportunities for crosssectional description of these types of relationships as well. Relating improvements in union quality to women’s employment requires longit udinal data. Fragile Families could support this, as could NLSY (which measures union quality annually). NSFH measures union quality, but the interviews are widely spaced. Nonetheless, as it collects general work histories covering the intervals between waves, it would also support this type of analysis. Finally, the special issues faced by single mothers could perhaps be addressed using the Fragile Families subsample of mothers who were no longer romantically involved with the father at the time of the birth of the child (n=562), or mothers who were no longer romantically involved with the father at the time of the 12-month follow-up (n=1377). The five data sets cited with regard to the first two topics would also allow analysis of the relationship between single mothers’ employment and union formation, with the same caveats. Theme 5: Men’s Economic Status Research issues: Calls to improve unskilled men’s economic prospects have been an important aspect of discussions of alternative approaches to marriage promotion. The expected importance of men’s economic status derives from the weight traditionally placed on men’s fulfilling the primary Abt Associates Inc. Chapter 11: Research Opportunities 127 breadwinner role when men and women specialized in different household roles. Recent research consistently in dicates that the male breadwinner role continues to matter: men’s economic status is positively related to their likelihood of getting and staying married. Better-off men also are somewhat more likely to start cohabiting than remaining single, and to marry if they are cohabiting. That said, analyses typically have found that men’s economic statuses do not explain very much of the variation in marriage at any given time, over time, or across racial and ethnic groups. The concept of men’s economic status is multi-dimensional and requires better specification than it typically has received (one promising approach emphasizes career development processes). For the most part, the absolute difference between husbands’ and wives’ resources has not been found to be as important as whether or not wives have resources: when they do not, their husbands’ income matters more. How steady increases in women’s economic status may be affecting the importance of men’s earnings for marriage and cohabitation is an important topic for further study. Studies of variation in responses within the general population of disadvantaged men are needed to gauge the nature and extent of economic improvement required for meaningful impacts on marriage prospects. Data needs: Studies of these issues require time-varying measures of men’s economic status, including educational attainment, school enrollment, job stability, occupation, and earnings. Similar measures on their female partners would help assess the effects of relative economic status. For single men, area level descriptors of potential mates are needed. Constructing the outcome variables requires full marriage and cohabitation histories. Consideration of cultural aspects requires measures of values and attitudes at different points in time. Analysis samples should include successive cohorts observed through their early 40s. Assessment of data sources: All of the data sources have information on the employment and earnings of men and of their partners, and most also have marriage and fertility histories. (Note that the relevant part of the NSFG is Cycle 6, Males; previous cycles interviewed women only.) Restricting the sources to those with at least approximate cohabitational histories and with longitudinal data, however, leaves only the NLSY, NSFH, and PSID. Fragile Families can be included for prospective analysis, i.e. for tracing the outcomes of the unions that produced the focal child as a function of the father’s economic status. NSFG Cycle 6 can perhaps also be considered, for an analysis based on men’s retrospective work history correlated with marital and (partial) cohabitational history—information is collected only on the current and first cohabiting partners. The most natural choices, however, are NLSY, NSFH, and PSID, which collect data on the economic status of men (and women) prior to union formation. Ellwood and Rodda (1990) and Oppenheimer et al. (1997) previously used the NLSY to study the effects of men’s economic status on marriage formation; Koball (1998) used Wave 1 of the NSFH to do so. Theme 6: The Meaning of Marriage and Cohabitation Research issues: Evidence of declines in the significance of marriage has inspired some in the marriage movement to view culture change as fundamental to the success of efforts to reinstitutionalize marriage. The vast majority of the population continues to see marriage as important, but the standards against which people are assessing the case for marriage appear to have changed. Whereas people formerly saw marriage as an inevitable and central aspect of life, now alternatives such as single parenting, cohabitation, and divorce have become acceptable. Their union decisions now give more weight to perceived short-run benefits and costs. Studies have found that blacks place more emphasis than whites on financial prerequisites for marriage. Although it is likely that such 128 Chapter 11: Research Opportunities Abt Associates Inc. priorities reflect blacks’ greater economic difficulties, we know little about how poverty affects the formation of values and expectations. Turning to the effects of values and expectations, a number of studies have found that values pertaining to marriage, divorce, and personal autonomy affect the formation and longevity of marriage and cohabitation. Social scientists suspect that many people may be setting the bar for marriage too high—placing expected benefits out of reach for many couples and thereby discouraging marriage and increasing marital strife and instability. Ethnographic research generally supports this thesis—especially among poor people —but there has been little quantitative verification. Finally, although we know much about the changing significance of marriage, there has been relatively little direct study of the meaning of cohabitation and how views of cohabitation affect experiences in both formal and informal unions. Data needs: Measures are needed for norms and values, perceived benefits and costs, and expected experiences within unions, with consistent measures for both marriage and cohabitation. Particular issues to be addressed include the meaning within each union type of childbearing, commitment, social status, financial security, and emotional well-being. Measures are also needed of other values that can affect relationships, including topics such as the importance of individualism, secularization, work orientation, and consumption aspirations. Longitudinal data are needed with repeated measurement of individual values and attitudes at different points in life course and at different stages in the evolution of relationships. Assessment of data sources: Of the nine data sources, five have some measures of values and attitudes. The best that can be done for measuring norms in this type of data is to aggregate attitudes and values from similar respondents. Fragile Families collects information from both partners on their general view of marriage, the importance of various factors for a successful marriage, and their views about the roles of men as fathers. These views can be linked to measures of union quality and subsequent union outcomes. The NLSY has only limited relevant information: attitudes towards men’s and women’s roles, and occupational plans. Women’s occupational plans could perhaps be linked to their union outcomes. NSAF has several items on attitudes towards single mothers and working mothers raising children. The cross-sectional nature of these data, however, implies that they could only be used for descriptive analysis rather than predicting outcomes from attitudes. Similarly, NSFG, though obtaining interesting measures of attitudes on family life and the rights and status of women, is handicapped by its cross-sectional nature. Analysis of the repeated cross-sections could however shed light on changes in norms over time. Finally, the NSFH offers the richest array of value and attitudinal measures, and follows the respondents through three interviews spanning 14 years. Cohabitors and other unmarried individuals (the latter up to age 35 only) are asked about perceived changes in their lives if they married, marital timing considerations, marriage considerations, and attitudes for and against cohabitation. In addition, cohabitors are asked about their plans for marriage and attitudes towards marriage. Abt Associates Inc. Chapter 11: Research Opportunities 129 Theme 7: Gender Role Expectations Research issues: An important subclass of expectations concerns norms about the appropriate roles of men and women in unions. In particular, the degree to which people hold traditional gender role expectations—i.e., with men specializing in market activities and women in home production—is thought to play an important role in moderating the effects of economic factors. As women have moved into the workforce, norms about gender roles within marriage have become increasingly egalitarian. Indirect evidence suggests that even more egalitarian norms apply to cohabitation. General population analyses find that more traditional role expectations predict entries to marriage, whereas more egalitarian values predict cohabitation. Studies tend also to find that gender role expectations have an important role in moderating the effects on union stability and quality of men and women’s economic statuses. Blacks hold more traditional views of appropriate gender roles than whites, further raising barriers to marriage associated with the poor earnings prospects of many black men. There are, however, indications that people will adjust their attitudes and behaviors, even while continuing to hold onto traditional values. Further research on the processes linking norms, va lues, attitudes, and behaviors could help to assess the potential for lowering the bar by fostering more positive views among men of home making and fathering roles. Although race-ethnicity differences have been well studied, it is unknown whether traditional views of gender roles are a more general concomitant of poverty, and how gender role expectations affect marriage prospects in low-income populations. Of particular interest is how policies increasing pressures on low-income women to work may be affecting marriage outcomes in segments of the low-income population with more and less traditional gender role expectations. Data needs: Measures of gender role expectations are needed along with information on broader values pertaining to egalitarianism, child bearing, material consumption, and work. Although information is needed from both men and women, it is even more useful to have this information from both members of a couple. Longitudinal data are needed to relate attitudes to subsequent union outcomes, and to trace evolution of attitudes. Assessment of data sources: Four of the nine data sources include some measures of gender role attitudes, from both men and women: Fragile Families, NLSY, NSFG and NSAF. (Surprisingly, although NSFH collects extensive attitudinal measures, none relate directly to gender roles). Only one of these, Fragile Families, collects parallel information from both members of the couple. Fragile Families includes items on attitudes towards appropriate roles of men and women and expectations on men’s roles as fathers. The NLSY measured respondents’ attitudes towards men’s and women’s roles three times, with a battery of eight items. NSAF includes two relevant items about attitudes towards working mothers. Because these latter data are cross-sectional, however, attitudes cannot be related to union outcomes. NSFG includes an 18-item battery on gender roles. These data are also crosssectional, but comparison of the various waves can shed light on changing norms and attitudes. Theme 8: Interaction Processes Research issues: Patterns of behavior and underlying processes characterizing couple interaction strongly predict long-term marital success. Findings from basic research on interaction have many important implicatio ns for prevention of relationship distress and treatment of distressed couples. Extensive observation of convenience samples of largely white, middle -class married samples shows that couples who exchange few positive behaviors relative to negative ones and who criticize or 130 Chapter 11: Research Opportunities Abt Associates Inc. express belligerence and contempt are especially likely to experience marital disruption. Patterns, such as negative reciprocity and “demand-withdrawal” behavior, have substantial potential for escalation. Cognitive factors (e.g., negative attributions), emotional expressions (e.g., negative affect), and physiological arousal (e.g., nervous system “flooding”) play important roles in determining the quality of interaction. To date, much of what we know is based on measurements in psychologists’ laboratories, and there has been little research on low-income couples. The rising importance of non-marital cohabitation, especially for low-income population, suggests an urgent need for comparative study of interaction patterns and underlying processes across more varied types of unions, and of the effects of differences in patterns of interaction on long-term relationship outcomes. Given how well early marital interaction predicts later outcomes, it would be helpful to understand better the processes at work during even earlier stages of relationship formation. Data needs: Research in this area requires measures of interaction behaviors, cognitive processes and inferences, affective expressions, and physiological correlates of observed behavior. Couples should be observed interacting in various natural situations in which these measures could be taken. Multidimensional measures of marital satisfaction are also needed. Assessment of data sources: None of the data sources examined here collects observational data on couples. Several sources—Fragile Families, NLSY, NSFH, SPD—do, however, collect more or less detailed information on union quality and satisfaction from samples that are nationally representative or approximately so, and that could be analyzed for a variety of disadvantaged groups. These data are useful for present purposes to the extent they can be linked to behaviors, cognitive processes, or affective expressions. The four data sets offer varying abilities to do so. Fragile Families includes information on joint activities, subjects of disagreement, rapport and trust, and frequency of certain positive and negative behaviors. It collects similar information from both members of the couple. The NSFH collects data on measures of perceived fairness of role allocation, time spent together, sexual frequency, areas of disagreement with partner, methods of coping with disagreement, and physical arguments. The NLSY allows union satisfaction to be linked to frequency of various interactions (calm discussion, laughing together, describing one’s day) and frequency of arguments on various subjects (e.g. chores, children, money). The SPD collects union quality data only on couples with children. (In fact, the items pertain to the relationship between the focal child’s two parents, regardless of whether they are living together.) For these couples, it is possible to relate reported satisfaction and consideration of separation in the last few months to physical arguments in the past year, as well as to disagreements about household tasks, money, spending time together, sex, in-laws, and children. For all of these data sets, these relevant measures are collected at multiple points in time, allowing prospective analysis of the influence of interactive processes on union outcomes. The collection of such measures in national data sets permits tests of differences in these relationships for various disadvantaged groups. Theme 9: Intrapersonal Influences on Interaction Research issues: Efforts to develop broader social-psychological theories have argued the need to understand the effects on couple interaction of enduring personal characteristics and of social and economic environments in which interaction occurs. Relatively stable background characteristics such as education, ethnicity, and early childhood experiences affect marital outcomes indirectly through their influence on expectations, values, and appraisals and, ultimately, on positive, negative, Abt Associates Inc. Chapter 11: Research Opportunities 131 and problem-solving behaviors during interaction. Personality characteristics such as neuroticism and depression also can interfere with cognitive, affective, and physiological responses needed to sustain positive interaction. Understanding the role of such factors is essential in preventing and treating relationship distress. Researchers have begun to study the direct and moderating effects of personal dispositions among middle -class married couples. There has been little research on disadvantaged couples or informal unions. The former, in particular, deserve attention because their lower levels of education and higher levels of depression, mental illness, and substance abuse suggest serious barriers to maintaining healthy relationships. Data needs: Key for analysis of this topic are measures of personality traits. Measures of union outcomes and processes are needed as well. Measures of background characteristics that may interact with the traits in affecting union outcomes and processes are also needed. Assessment of data source s: The four data sources previously cited include varying information on respondents’ personality traits. (We refrain from repeating the description of the measures of union outcomes and processes available in these data sets.) Fragile Families collects information from both partners on depression and anxiety. NLSY has topical modules that address self-esteem (1979, 1980, 1987), sociability (1985), locus of control (1992), and depression (1992, 1994). NSFH includes a global measure of happiness and a 12-item depression scale. SPD measures depression only. These four data sets also include information on respondents’ backgrounds beyond the usual demographics, education, and work histories that might interact with personality traits in their influence on union behaviors. Fragile Families ascertains the birthplace and educational attainment of the respondent’s parents, and current relationship between the respondent and his or her parents, for both members of the couple. NLSY collected detailed information on respondents’ family background, including home life at age 14 (rural/urban, adults in household, their work characteristics and occupations, receipt of newspapers and magazines, possession of a library card), and living arrangements from birth to age 18. NSFH ascertains the respondent’s parents’ occupation and education, and receipt of public assistance during the respondent’s childhood. The SIPP modules for the SPD include migration history and recipiency history. Theme 10: Contextual Influences on Interaction Research issues: Many environmental conditions—including virtually all of the demographic, economic, and socio-cultural factors assessed for this review—can influence couple interaction and moderate the effects of interaction on union outcomes. Understanding these linkages is critical to knowing when and how environmental changes might be beneficial and identifying strategies for helping couples function better in challenging environments. Cultural differences in norms and values governing interpersonal behavior may help to explain subgroup variation in approaches to problem-solving communication. Another important contextual influence is exposure to stressful events. In particular, financial stress has been observed in prospective studies to have strong negative effects on marital quality and stability. Both chronic and acute stresses—the latter arising from events such as a job loss, car breakdown, or eviction—can wreak havoc on the effortful cognitive transformations required to sustain positive couple interaction. Understanding why some couples are more resilient than others in the face of stress may lead to clinically useful insights. There are strong indications that spouses who exhibit supportive behaviors when their partners lose a job greatly 132 Chapter 11: Research Opportunities Abt Associates Inc. ameliorate the negative effects of stress on their relationships. Most of the research on financial stress and job loss to date has pertained to outcomes for convenience samples of middle -class white married couples. Hence, although the coincidence of multiple sources of stress is likely to be especially damaging for poor peoples’ relationships, we know little about such effects. Furthermore, virtually all of the research on this topic has pertained to engaged or married couples. There is a pressing need for more research on interaction within informal unions and on processes during early stages of relationship formation. Data needs: The data needs are the same measures of union outcomes and processes as for the preceding two themes, with the addition of measures of the context of couples’ varying situations. Four dimensions of importance are childbearing and the presence of children; stressful life events and crises, especially financial stressors; social norms; and personal characteristics that can moderate or exacerbate stresses on a relationship. Such analyses require detail on the timing of relevant context, process, and outcome variables to assign correct causal influence. Assessment of data sources: Of the four data sources mentioned previously, Fragile Families and SPD cannot be used to study the moderating effect of children, as all Fragile Family sample members have a new baby, and the SPD union quality questions are asked only of couples with children. Both these data sets as well as NLSY and NSFH have information on employment and income relative to household size, and all but Fragile Families ask about assets and debts. Variations in social norms are probably best proxied in these data sets by respondents’ religion, whic h is measured in all but the SPD; and perhaps by race and ethnicity, which are measured in all of them. All four sources measure respondents’ health and education, which may moderate stresses on a relationship. Marriage attitudes, which may do likewise, are measured in Fragile Families. Substance abuse is captured in Fragile Families and SPD, and depression in all four (in a topical module for NLSY). Abt Associates Inc. Chapter 11: Research Opportunities 133 References Ellwood, D. T. & Jencks, C. (2001). The growing differences in family structure: What do we know? Where do we look for answers? Unpublished manuscript, Harvard University. Ellwood, D.T. & Rodda, D. (1990). The hazards of work and marriage: The influence of employment on marriage. Harvard University (mimeo). Fein, D.J., Burstein, N., Fein, G.G., & Lindberg, L.D. (2003). The Determinants of Marriage and Cohabitation among Disadvantaged Americans: Research Findings and Needs. Bethesda, MD: Abt Associates Inc. Federal Interagency Forum on Child and Family Statistics (2001). Counting Couples: Improving Marriage, Divorce, Remarriage, and Cohabitation Data in the Federal Statistical System, Conference Handout. Bethesda, MD: The National Institutes of Health. Federal Interagency Forum on Child and Family Statistics (2002). Counting Couples: Improving Marriage, Divorce, Remarriage, and Cohabitation Data in the Federal Statistical System, Highlights from a National Workshop December 13 and 14, 2001. Bethesda, MD: The National Institutes of Health. Hoffman, S.D., Duncan, G.J., & Mincy, R. (1991), Marriage and welfare use among young women, paper presented at Population Association of America. Koball, H. (1998). Have African American men become less committed to marriage? Explaining the twentieth century racial cross-over in men's marriage timing. Demography, 35(2), 251-258. Lichter, D. T., & Graefe, D. R. (2001). Finding a mate? The marital and cohabitation histories of unwed mothers. In L. L. Wu & B. Wolfe (Eds.), Out of Wedlock: Causes and Consequences of Nonmarital Fertility (pp. 317-343). New York: Russell Sage Foundation. Lichter, D.T., McLaughlin, D.K., Kephart, G., and Landry, D.J. (1992). Race and the retreat from marriage: A shortage of marriageable men? American Sociological Review, 57, 781-799. McLanahan, S. & Sanderfur, G. (1994). Growing Up With A Single Parent: What Hurts, What Helps. Cambridge, MA: Harvard University Press. Oppenheimer, V. K., Kalmijn, M., & Lim, N. (1997). Men's career development and marriage timing during a period of rising inequality. Demography, 34(3), 311-330. Zill, N. & Daly, M. (1993). Researching the Family: A Guide to Survey and Statistical Data on U.S. Families. Washington, DC: Child Trends, Inc. Abt Associates Inc. References 135 Appendix A: Research Opportunities by Theme Exhibit A.1 Research Opportunities for Teen and Non-Marital Childbearing Data Set 1. Current Population Survey (CPS) Fragile Families and Child Wellbeing Survey (FF) Measures Have marriage and fertility history, but no cohabitation history. No measures of union quality. Rich information on union quality from both partners. Attitudinal data from both partners on views of marriage and parenthood. No prior marriage, fertility, or cohabitation history. Biological and stepchildren not distinguished. Have marriage and fertility history; cohabitation information since 1990. Annual measures of union quality. Sample and Structure Effectively cross-sectional. 2. Sample comprises couples with a new infant. Follow-up at one and three years. 3. National Longitudinal Survey of Youth (NLSY) Full nationally representative sample. 4. 5. National Survey of American Families (NSAF) National Survey of Family Growth (NSFG) No measures of marital history or relationship quality. Full history of marriage, cohabitation, fertility. Minimal data on quality (reasons for separations). Cannot clearly identify biological parents of each child. Full history of marriage, cohabitation, fertility. Rich information on union quality from both partners. Full history of marriage and fertility, crude history of cohabitation. No information on union quality. Full history of marriage and fertility, but not cohabitation. No information on union quality. Full history of marriage and fertility. Cohabitation can be traced crudely over life of the panel. Several items on union quality. Cross-sectional, multiple waves. Cross-sectional, multiple waves. 6. National Study of Families and Households (NSFH) Sample members are interviewed at points in time spanning 14 years. Includes all household types. Nationally representative, panel. Supports analysis of sibling pairs. Nationally representative, panel. 7. Panel Study of Income Dynamics (PSID) Survey of Income and Program Participation (SIPP) Survey of Program Dynamics (SPD) 8. 9. Nationally representative, panel. Abt Associates Inc. Appendix A: Research Opportunities by Theme A-1 Exhibit A.2 Research Opportunities for Transition to Parenthood Data Set 1. 2. Current Population Survey (CPS) Fragile Families and Child Wellbeing Survey (FF) Measures No measures of union quality. Rich information on union quality from both partners, including couple’s relationship with each other and each member’s relationship with focal child. Annual measures of union quality, including satisfaction, frequency of various interactions, and frequency of arguments on various subjects. No measures of relationship quality. Minimal data on quality (reasons for separations). Retrospective data on wantedness and other circumstances surrounding each pregnancy. Rich information on union quality from both partners. Information on relationship with children, stepchildren, biological children living elsewhere. Good contextual data, but not focused on pregnancy. No information on union quality. No information on union quality. Several items on union quality. Rich contextual data, but not focused on pregnancy. Sample and Structure Effectively cross-sectional. Sample comprises couples with a new infant. Follow-up at one and three years. 3. National Longitudinal Survey of Youth (NLSY) Full nationally representative sample. Includes cohabitors, single mothers, stepfamilies. Cross-sectional, multiple waves. Cross-sectional, multiple waves. 4. 5. National Survey of American Families (NSAF) National Survey of Family Growth (NSFG) 6. National Study of Families and Households (NSFH) Sample members are interviewed at three points in time spanning 14 years. Includes all household types. 7. 8. 9. Panel Study of Income Dynamics (PSID) Survey of Income and Program Participation (SIPP) Survey of Program Dynamics (SPD) Nationally representative, panel. Nationally representative, panel. Nationally representative, panel. Includes all household types. A-2 Appendix A: Research Opportunities by Theme Abt Associates Inc. Exhibit A.3 Research Opportunities for Effects of Cohabitation on Marriage Data Set 1. 2. Current Population Survey (CPS) Fragile Families and Child Wellbeing Survey (FF) Measures No measures of cohabitation history or characteristics of relationship. Rich information on union quality and characteristics, relationship with family and friends, attitudes towards marriage and parenting. Annual measures of union quality. No measures of attitudes towards marriage, etc. No measures of cohabitation history or characteristics of relationship. Complete cohabitation history. Attitudes on family life, rights and status of women. Cycle 6 includes attitudes on marriage and cohabitation from both men and women. Rich information on union quality from both partners. Attitudes towards marriage and cohabitation, marital timing. No information on union quality or attitudes. No information on characteristics of relationship or attitudes. No information on characteristics of relationship or attitudes. Sample and Structure Effectively cross-sectional. Sample comprises couples with a new infant. Follow-up at one and three years. Single cohort. Full nationally representative sample. Single cohort. Cross-sectional, multiple waves. Cross-sectional, multiple waves. 3. National Longitudinal Survey of Youth (NLSY) National Survey of American Families (NSAF) National Survey of Family Growth (NSFG) 4. 5. 6. National Study of Families and Households (NSFH) Single cohort. Sample members are interviewed at three points in time spanning 14 years. Nationally representative, panel. Nationally representative, panel. Nationally representative, panel. 7. 8. 9. Panel Study of Income Dynamics (PSID) Survey of Income and Program Participation (SIPP) Survey of Program Dynamics (SPD) Abt Associates Inc. Appendix A: Research Opportunities by Theme A-3 Exhibit A.4 Research Opportunities for Surge of Low-Income Women Into Jobs Data Set 1. Current Population Survey (CPS) Measures Good measures of employment including occupation, sector, type of business, usual activities. No measures of union quality. Standard information on employment, plus off-the-books work and illegal activities for money. Measures of union quality. Detailed data on employment, plus topical modules on work world, attitudes toward work, self-esteem, and vocational ability. Annual measures of union quality. Good information on earnings and employment, including type of employer, occupation, hours, work schedule, child care. No measures of union quality. Information on occupation, activities and duties, type of business, hours, work schedule, earnings, child care arrangements. Work history. Minimal information on union quality (reasons for separations). Information on occupation, hours, earnings, work schedule, child care. Work history. Good data on union quality. Information on union membership, occupation, duties, type of business, hours worked. Work history, vocational training. Achievement motivation (1972). No data on union quality. Information on employer size and location, occupation, industry, work activities and duties, hours. Topical modules on work schedule, education and training. No measures of union quality. Type of employment, hours worked, occupation, duties. Employment history. No measures of union quality. Sample and Structure Effectively cross-sectional. Have coverage of 16 months but not if people move. Sample comprises couples with a new infant. Follow-up at one and three years. Full nationally representative sample. But youngest members are aged 36 by 2000. Cross-sectional, multiple waves. 2. Fragile Families and Child Wellbeing Survey (FF) 3. National Longitudinal Survey of Youth (NLSY) 4. National Survey of American Families (NSAF) 5. National Survey of Family Growth (NSFG) Cross-sectional, multiple waves. 6. National Study of Families and Households (NSFH) Sample members are interviewed at three points in time spanning 14 years. Nationally representative, panel. 7. Panel Study of Income Dynamics (PSID) 8. Survey of Income and Program Participation (SIPP) Nationally representative, panel. 9. Survey of Program Dynamics (SPD) Nationally representative, panel. A-4 Appendix A: Research Opportunities by Theme Abt Associates Inc. Exhibit A.5 Research Opportunities for Men’s Economic Status Data Set 1. Current Population Survey (CPS) Measures Good measures of employment including occupation, sector, type of business, usual activities. Marriage and fertility supplement. No cohabitational history. Standard information on employment, plus off-the-books work and illegal activities for money. Job skill training. No prior marital, cohabitational, fertility history. Some job history information. Detailed data on employment, plus topical modules on work world, attitudes toward work, self-esteem, and vocational ability. Employment history. Marital, cohabitational, fertility history. Good information on earnings and employment, including type of employer, occupation, hours, work schedule. No marital, cohabitational, fertility, employment history. Information on current employment and earnings. Complete marital and fertility history, partial cohabitational history. Some work history. Information on current employment and earnings. Marital, cohabitational, fertility, and work histories. Information on current employment and earnings. Marital and fertility history. Approximate cohabitational his tory. Work history, vocational training. Employment and earnings. Marital and fertility history, no cohabitational history. Work history, education and training. Employment and earnings. Marital and fertility history, no cohabitational history. Work history, education and training. Sample and Structure Effectively cross-sectional. Have coverage of 16 months but not if people move. Sample comprises couples with a new infant. Followup at one and three years. 2. Fragile Families and Child Wellbeing Survey (FF) 3. National Longitudinal Survey of Youth (NLSY) Full nationally representative sample. But youngest members are aged 36 by 2000. Cross-sectional, multiple waves. 4. National Survey of American Families (NSAF) 5. National Survey of Family Growth (NSFG) Cross-sectional, multiple waves. 6. National Study of Families and Households (NSFH) Panel Study of Income Dynamics (PSID) Sample members are interviewed at three points in time spanning 14 years. Nationally representative, panel. 7. 8. Survey of Income and Program Participation (SIPP) Survey of Program Dynamics (SPD) Nationally representative, panel. Nationally representative, panel. 9. Abt Associates Inc. Appendix A: Research Opportunities by Theme A-5 Exhibit A.6 Research Opportunities for the Meaning of Marriage and Cohabitation Data Set 1. 2. Current Population Survey (CPS) Fragile Families and Child Wellbeing Survey (FF) Measures No measures of attitudes or values. Information on attitudes and values from both partners. General view of marriage, keys to a successful marriage, roles of men and women. Topical modules on attitudes towards men’s and women’s roles (1979, 1982, 1987), expectations for marital timing (1979, 1981), fertility (most years), occupational plans (19791984). Attitudes on welfare, raising children vis-à-vis work and marriage. Plans to marry, attitudes on family life, rights and status of women, expectations for fertility, desired family size. Fertility intentions, fertility considerations, evaluation of roles; for non-married, non-cohabitors age 35 and younger, and cohabitors: perceived changes in life if married, marital timing considerations, marriage considerations, attitudes for and against cohabitation; for cohabitors, plans for marriage and attitudes towards marriage. No measures of attitudes or values. No measures of attitudes or values. No measures of attitudes or values. Sample and Structure Effectively cross-sectional. Sample comprises couples with a new infant. Follow-up at one and three years. Single cohort. Full nationally representative sample. Single cohort. 3. National Longitudinal Survey of Youth (NLSY) 4. 5. National Survey of American Families (NSAF) National Survey of Family Growth (NSFG) Cross-sectional, multiple waves. Cross-sectional, multiple waves. 6. National Study of Families and Households (NSFH) Single cohort. Sample members are interviewed at three points in time spanning 14 years. 7. 8. 9. Panel Study of Income Dynamics (PSID) Survey of Income and Program Participation (SIPP) Survey of Program Dynamics (SPD) Nationally representative, panel. Nationally representative, panel. Nationally representative, panel. A-6 Appendix A: Research Opportunities by Theme Abt Associates Inc. Exhibit A.7 Research Opportunities for Gender Role Expectations Data Set 1. 2. Current Population Survey (CPS) Fragile Families and Child Wellbeing Survey (FF) National Longitudinal Survey of Youth (NLSY) National Survey of American Families (NSAF) National Survey of Family Growth (NSFG) National Study of Families and Households (NSFH) Panel Study of Income Dynamics (PSID) Survey of Income and Program Participation (SIPP) Survey of Program Dynamics (SPD) Measures No measures of attitudes. Information from both partners on roles of men and women, roles of fathers. Topical modules on attitudes towards men’s and women’s roles (1979, 1982, 1987). Attitudes on raising children vis-àvis work and marriage. Attitudes on family life, rights and status of women. No measures of gender role attitudes. No measures of attitudes . No measures of attitudes. No measures of attitudes. Sample and Structure Effectively cross-sectional. Sample comprises couples with a new infant. Follow-up at one and three years. Single cohort. Full nationally representative sample. Single cohort. Cross-sectional, multiple waves. Cross-sectional, multiple waves. Single cohort. Sample members are interviewed at three points in time spanning 14 years. Nationally representative, panel. Nationally representative, panel. Nationally representative, panel. 3. 4. 5. 6. 7. 8. 9. Abt Associates Inc. Appendix A: Research Opportunities by Theme A-7 Exhibit A.8 Research Opportunities for Interaction Processes Data Set 1. 2. Current Population Survey (CPS) Fragile Families and Child Wellbeing Survey (FF) Measures No measures of union quality. Information on union quality from both partners. Joint activities, subjects of disagreement, rapport and trust, frequency of certain positive and negative behaviors. Annual measures of union quality, interactions, arguments. No measures of union quality. Minimal data on union quality (reasons for separation). Rich information on union quality from both partners. Perceived fairness of role allocation, time spent together, sexual frequency, areas of disagreement with partner, methods of coping with disagreement, physical arguments No information on union quality. No information on union quality. For couples with children only: annual items on union quality, consideration of separation, physical arguments. Areas of disagreement with other parent. Sample and Structure Effectively cross-sectional. Sample comprises couples with a new infant. Follow-up at one and three years. Single cohort. Full nationally representative sample. Single cohort. Cross-sectional, multiple waves. Cross-sectional, multiple waves. Single cohort. Sample members are interviewed at three points in time spanning 14 years. 3. 4. 5. 6. National Longitudinal Survey of Youth (NLSY) National Survey of American Families (NSAF) National Survey of Family Growth (NSFG) National Study of Families and Households (NSFH) 7. 8. 9. Panel Study of Income Dynamics (PSID) Survey of Income and Program Participation (SIPP) Survey of Program Dynamics (SPD) Nationally representative, panel. Nationally representative, panel. Nationally representative, panel. A-8 Appendix A: Research Opportunities by Theme Abt Associates Inc. Exhibit A.9 Research Opportunities for Intrapersonal Influences on Interaction Data Set 1. 2. Current Population Survey (CPS) Fragile Families and Child Wellbeing Survey (FF) Measures No measures of union quality. Information on union quality from both partners. Joint activities, subjects of disagreement, rapport and trust, frequency of certain behaviors, abuse. Depression, anxiety. Annual measures of union quality, interactions, arguments Topical modules on self-esteem, sociability, locus of control, depression. No measures of union quality. Minimal data on union quality (reasons for separation). Rich information on union quality from both partners. Perceived fairness of role allocation, time spent together, sexual frequency, areas of disagreement with partner, methods of coping with disagreement, physical arguments. Global happiness, depression (12 items). No information on union quality. No information on union quality. For couples with children, annual items on union quality, consideration of separation, physical arguments. Areas of disagreement with the other parent. Depression. Sample and Structure Effectively cross-sectional. Sample comprises couples with a new infant. Follow-up at one and three years. Single cohort. Full nationally representative sample. Single cohort. 3. National Longitudinal Survey of Youth (NLSY) 4. 5. 6. National Survey of American Families (NSAF) National Survey of Family Growth (NSFG) National Study of Families and Households (NSFH) Cross-sectional, multiple waves. Cross-sectional, multiple waves. Single cohort. Sample members are interviewed at three points in time spanning 14 years. 7. 8. 9. Panel Study of Income Dynamics (PSID) Survey of Income and Program Participation (SIPP) Survey of Program Dynamics (SPD) Nationally representative, panel. Nationally representative, panel. Nationally representative, panel. Abt Associates Inc. Appendix A: Research Opportunities by Theme A-9 Exhibit A.10 Research Opportunities for Contextual Influences on Interaction Data Set 1. 2. Current Population Survey (CPS) Fragile Families and Child Wellbeing Survey (FF) Measures No measures of union quality. Information on union quality from both partners. Joint activities, subjects of disagreement, rapport and trust, frequency of certain behaviors, abuse. All couples have children. Data on employment, income, substance abuse, marriage attitudes, religion, education, health, depression. Annual measures of union quality, interactions, arguments. Data on children, employment, income, assets and debts, education, health; depression (topical module). No measures of union quality. Minimal data on union quality (reasons for separation). Rich information on union quality from both partners. Perceived fairness of role allocation, time spent together, sexual frequency, areas of disagreement with partner, methods of coping with disagreement, physical arguments. Data on children, employment, income, assets and debts, religion, education, health, depression. No information on union quality. No information on union quality. For couples with children, annual items on union quality, consideration of separation, physical arguments. Areas of disagreement with the other parent. Data on employment, income, assets and debts, substance abuse, education, depression. Sample and Structure Effectively cross-sectional. Sample comprises couples with a new infant. Follow-up at one and three years. Single cohort. 3. National Longitudinal Survey of Youth (NLSY) Full nationally representative sample. Single cohort. 4. 5. 6. National Survey of American Families (NSAF) National Survey of Family Growth (NSFG) National Study of Families and Households (NSFH) Cross-sectional, multiple waves. Cross-sectional, multiple waves. Single cohort. Sample members are interviewed at three points in time spanning 14 years. 7. 8. 9. Panel Study of Income Dynamics (PSID) Survey of Income and Program Participation (SIPP) Survey of Program Dynamics (SPD) Nationally representative, panel. Nationally representative, panel. Nationally representative, panel. A-10 Appendix A: Research Opportunities by Theme Abt Associates Inc.